(Received for publication, August 6, 1996, and in revised form, December 13, 1996)
From the Department of Molecular Physiology and Biological Physics, University of Virginia Health Sciences Center, Charlottesville, Virginia 22908
The goal of the present study was to determine
the molecular mechanism whereby transforming growth factor (TGF
)
increases smooth muscle (SM)
-actin expression. Confluent,
growth-arrested rat aortic smooth muscle cells (SMC) were transiently
transfected with various SM
-actin promoter/chloramphenicol
acetyltransferase deletion mutants and stimulated with TGF
(2.5 ng/ml). Results demonstrated that the first 125 base pairs of the SM
-actin promoter were sufficient to confer TGF
responsiveness.
Three cis elements were shown to be required for TGF
inducibility: two highly conserved CArG boxes, designated A (
62) and
B (
112) and a novel TGF
control element (TCE) (
42). Mutation of
any one of these elements completely abolished TGF
-induced reporter
activity. Results of electrophoretic mobility shift assays demonstrated
that nuclear extracts from TGF
-treated SMC enhanced binding activity
of serum response factor to the CArG elements and binding of an as yet
unidentified factor to the TCE. Northern analysis showed that TGF
also stimulated transcription of two other SM (SM myosin heavy chain)
differentiation marker genes, SM myosin heavy chain and h1
calponin, whose promoters also contained a TCE-like element. In
summary, we identified a TGF
response element in the SM
-actin
promoter that may contribute to coordinate regulation of expression of
multiple cell-type specific proteins during SMC differentiation.
Restenosis after balloon angioplasty results from abnormal
proliferation of phenotypically modulated vascular smooth muscle cells
(SMC)1 that synthesize large amounts of
extracellular matrix (1). A variety of growth factors have been shown
to play a role in the development of restenotic lesions including
transforming growth factor 1 (TGF
) (2, 3). The expression of
TGF
mRNA and protein in the arterial wall is increased following
balloon injury in rats (3) and overexpression of TGF
by gene
transfer into normal arteries results in substantial extracellular
matrix production accompanied by intimal and medial hyperplasia (4).
Additionally, it has been shown that intravenous administration of
neutralizing anti-TGF
antibodies significantly reduced intimal
lesion size following carotid injury in the rat (5). The precise role
of TGF
in lesion development, however, is not known, and is
confounded by observations that TGF
can have very diverse functions
in cultured SMC (6, 7). In particular, TGF
can either stimulate or inhibit cell proliferation depending on the cell line and culture conditions employed (8, 9). In addition to its effects on cell growth,
TGF
has also been implicated in differentiation of SMC and other
mesenchymal-derived cells during development (10). For example, TGF
treatment of human SMC (11), granulation tissue myofibroblasts (12),
and pericytes (13), has been shown to increase expression of SM
-actin, a marker of differentiated SMC (14). Whereas SM
-actin is
also expressed in cardiac and skeletal muscle during development (15),
and in fibroblasts during wound repair (16), its expression is highly
specific for SMC under normal circumstances in adult animals (14). It is the single most abundant protein in SMC, and is required for the
principle function of mature SMC, contraction (14). Of particular interest, a recent study by Shah et al. (17), demonstrated
that TGF
stimulated differentiation of neural crest cells into SMC or SMC-like cells based on morphological criteria and induction of SM
-actin and calponin. However, the molecular mechanisms whereby
TGF
up-regulates SM
-actin expression are not known, nor is it
known whether TGF
alters expression of other SMC differentiation marker proteins.
SM -actin expression has been shown to be governed by a complex
interplay of both positive and negative acting cis elements (18-21) depending on the cell context studied. The SM
-actin
promoter contains several highly conserved cis elements
including CArG elements. CArG elements are also found in the promoters
of skeletal and cardiac
-actin genes as well as many other
muscle-specific genes and have been shown to direct developmental and
tissue-specific expression of these genes (22-28). Studies in our
laboratory demonstrated that two CArG elements designated A and B, that
are located within the first 125 bp of the rat SM
-actin promoter
were required for basal and cell-specific expression (29). These latter
results are of particular interest since MacLellan et al.
(30) previously demonstrated that CArG elements were also important for
TGF
inducibility of the skeletal
-actin gene in cardiac myocytes,
thus suggesting a possible mechanism whereby TGF
might alter
expression of the SM
-actin gene in vascular SMC.
The aims of the present study were to address the following questions:
1) what sequences impart TGF responsiveness to the SM
-actin
promoter? 2) Do the identified sites confer both basal and TGF
inducible expression, or are these sequences unique to TGF
induction? 3) What trans-acting factors interact with these sites? 4) Is TGF
inducibility restricted to the SM
-actin gene or
does TGF
also stimulate expression of other SM differentiation marker genes including SM MHC and h1 calponin?
The
generation of various truncated SM -actin promoter/CAT reporter
constructs, including the CArG A and B mutants have been previously
reported (29). Site-directed mutagenesis of the TGF
control element
(TCE)-like sequence spanning from
48 to
57 (TCE wild type:
5
-ATGAGG-3
) was performed
using the Altered Sites in vitro mutagenesis system
(Promega) according to the manufacturer's recommendations. Three
separate mutants of this putative TGF
response element were made:
TCE mut 1: 5
-ATGGGAGG-3
; TCE mut 2:
5
-GAGTGAGG-3
; TCE mut 3:
5
-AGAGG-3
. The mutated fragments were polymerase chain reaction amplified and subcloned into
pCAT-Basic (Promega). The sequence was verified by the Sanger dideoxy-sequencing procedure (31) using a Sequenase kit (U. S. Biochemical Corp.).
All promoter-CAT plasmid DNAs used for transfections were prepared using an alkaline lysis procedure (32) followed by banding on two successive ethidium bromide-cesium chloride gradients. Multiple independent plasmid preparations were tested for each construct.
Cell Culture, Transient Transfections, and Reporter Gene AssaysRat SMC were isolated from thoracic aorta and cultured as
described previously (33). Cells were plated at a density of 3 × 103/cm2, grown to confluency in 10% serum
containing medium, and then growth-arrested for 4 days in serum-free
medium (34) prior to stimulation with TGF (human TGF
1 from R&D
Systems, 2.5 ng/ml) diluted with vehicle (4 mM HCl, bovine
serum albumin, 1 mg/ml). Control cultures were treated with vehicle
only. Cells used for the experiments described were between the 8th and
the 22nd passage. SMC that are growth-arrested in this fashion express
multiple SMC differentiation marker proteins including SM
-actin, SM
MHC, h-caldesmon, h1 calponin, SM
-tropomyosin, and SM
myosin light chain (MLC20)
(35-37).2
Confluent, growth-arrested SMC in 6-well plates were transiently
transfected (in triplicates) with 5 µg of DNA using the transfection reagent DOTAP (Boehringer Mannheim) according to the manufacturer's recommendations. After an incubation period of 12-14 h, the medium was
replaced with fresh serum-free medium and TGF (2.5 ng/ml) or vehicle
were added. Cells were harvested 72 h later by scraping into
ice-cold buffer A (15 mM Tris, pH 8.0, 60 mM
KCl, 15 mM NaCl, 2 mM EDTA, 0.15 mM
spermine tetrahydrochloride, 1 mM dithiothreitol) (38).
Cell lysates were prepared by four freeze-thaws, followed by 10 min
heat inactivation at 65 °C; 95-µl aliquots of each cell extract
were assayed for CAT activity by enzymatic butyrylation of tritiated
chloramphenicol (DuPont NEN) (39). CAT activities were normalized as
described previously (29). Experiments were repeated two to six times
and relative CAT activity data were expressed as the mean ± S.D.
unless otherwise noted.
Total RNA was isolated using TRI REAGENTTM
(Molecular Research Center, Cincinnati, OH) according to the
manufacturer's recommendations. Extracted RNA was dissolved in sterile
water and stored at 70 °C until use. RNA concentration was
measured spectrophoretically. The probe used for SM
-actin detection
consisted of a 512-bp EcoRI fragment that encoded for amino
acids 202 to 374 of the human skeletal
-actin cDNA (a gift
kindly provided by Drs. Gunning and Kedes, Veterans' Administration
Medical Center, Palo Alto, CA). For detection of SM MHC transcripts, a
373-bp cDNA probe corresponding to amino acids 1659 to 1929 of the
SM2 isoform of SM MHC (kindly provided by Drs. P. Babij and M. Periasamy) was used. This cDNA probe recognizes both the SM1 and
SM2 isoforms of SM MHC but not nonmuscle MHC (40, 41). H1
calponin transcripts were identified by a cDNA probe corresponding
to the coding region spanning nucleotides 144 to 715 (kindly provided
by M. Parmacek, University of Chicago). All cDNA probes were
isolated from plasmid vector DNA by appropriate restriction enzyme
digestions and gel purified (Bio-Rad, Prep-A-Gene). Probes were labeled
with [
-32P]dCTP (DuPont NEN) by random primer
extension (Prime-It, Stratagene). For Northern analysis, 10 or 15 µg
of total RNA was diluted in loading buffer consisting of 0.2 M MOPS, 0.05 M sodium acetate, 0.01 M EDTA, 4% formaldehyde, and 65% formamide, denatured by treating for 10 min at 65 °C, and subsequently resolved on a 1.2% agarose gel containing 6.1% formaldehyde and 10% loading buffer. Capillary transfer of RNA to a nylon membrane (Micron Separating, Westboro, MA) was carried out overnight in 10 × saline/sodium/phosphate/EDTA buffer (2 M EDTA, 20 M NaH2PO4 H2O, 298 M
NaCl). Blots were air dried, exposed to an UV transilluminator for 1.5 min, and baked for 2 h under vacuum at 80 °C.
Hybridization to cDNA probes and subsequent washes were carried out
at 65 °C as described previously by Church and Gilbert (42). Blots
were then dried, quantified using an Image Quant PhosphorImager, and
subsequently exposed to Kodak X-Omat K film at 70 °C. A 5.8-kb
EcoRI human cDNA fragment for 18 S rRNA was released
from pBR322 and used in Northern analysis for standardization of RNA
loading and transfer (43).
Cell lysates were prepared from
confluent, growth-arrested SMC cultures stimulated with TGF (2.5 ng/ml) or vehicle for 4 h. Briefly, cells were rinsed with
phosphate-buffered saline, scraped into 0.6 ml of ice-cold RIPA buffer
(phosphate-buffered saline, 1% Nonidet P-40, 0.5% sodium
deoxycholate, 0.1% SDS) plus protease inhibitors (10 mg/ml
phenylmethylsulfonyl fluoride, 30 µg/ml aprotinin, 100 mM
sodium orthovanadate), and passed through a 21-gauge needle for several
times. Cell lysates were then incubated on ice for 30 min and
microfuged for 20 min at 4 °C (protocol provided by Santa Cruz).
Sample loading was normalized to DNA content determined with a DNA
fluorometer (Hoefer Scientific, San Francisco). 600 ng of DNA was
loaded per well on a 7.5% SDS-PAGE Mini-Protean gel (Bio-Rad). The
proteins were transferred onto a polyvinylidene difluoride membrane at
100 V for 1.5 h. Blocking of the membrane and probing with
appropriate antibodies were performed according to the ECL Western
blotting protocol from Amersham and Life Science. Affinity-purified
rabbit polyclonal SRF antibodies (Santa Cruz), raised against a peptide
corresponding to SRF amino acids 486 to 505, were used as primary
antibodies at a concentration of 1 µg/ml.
Crude nuclear extracts were prepared by the
method of Dignam et al. (44) using confluent,
growth-arrested SMC stimulated with TGF or vehicle for 4 h.
Protein concentrations were measured by the Bradford assay (Bio-Rad).
Probes for EMSA were obtained by end labeling 20 µM
single-stranded oligonucleotides with 150 µCi of
[
-32P]ATP (6000 Ci/mmol) and T4 polynucleotide kinase.
Labeled single-stranded oligonucleotides were annealed and
unincorporated nucleotides were removed using NucTrap Push columns
(Stratagene) as recommended by the manufacturer. The following
nucleotides were used either as a probe or as cold competitors (only
sense strand shown): TCE, 5
-GAAGCGAGTGGGAGGGGAT-3
; TCE mut 2, 5
-GAAGCGAGTGTTAGGGGAT-3
; Sp1, 5
-GATCGATCGGGGCGGGGCGATC-3
; AP-1,
5
-CTAGTGATGAGTCAGCCGGATC-3
. The 20-µl binding reaction contained
~20,000 cpm labeled probe, 5 µg of nuclear extracts in Dignam
buffer D, 20 mM Hepes, pH 7.9, 50 mM KCl, 4 mM MgCl2, 0.2 mM EDTA, 0.5 mM dithiothreitol, 15% glycerol, 2.5% Nonidet P-40, 0.5 µg of poly(dA-dT) (Sigma), and cold competitor oligonucleotides where
indicated. Recombinant human Sp1 was added at a concentration of 9 fpU
per reaction (1 fpU defined as the amount of protein required to give
full protection against DNase I of Sp1 sites within SV 40 early
promoter (Promega)). All binding reactions were incubated for 20 min at
room temperature. For Sp1 or Sp3 supershift assays, antibodies (2 µg/reaction) were added to the binding reaction and incubated for 20 min at room temperature. The radiolabeled DNA was subsequently added to
the binding reaction and incubated for additional 20 min at room
temperature. Protein-DNA complexes were resolved on a 4.5%
polyacrylamide gel (30:1 acrylamide/bis-acrylamide (Bio-Rad)) and
electrophoresed at 170 V in 0.5 × TBE (45 mM Tris
borate, 1 mM EDTA). The gels were then dried and subjected
to autoradiography at
70 °C.
In vitro synthesis of
SRF was performed using a TNT coupled reticulocyte lysate translation
system (Promega) and using the human SRF cDNA clone pT7ATG (45)
as a template.
To
examine the effects of TGF on SM
-actin mRNA expression, rat
aortic SMC were grown to confluency, growth-arrested in SFM, and
treated with 2.5 ng/ml TGF
or vehicle. Total RNA was extracted from
these cultures at the times indicated and subjected to Northern blot
analysis for SM
-actin (Fig. 1). SM
-actin mRNA levels peaked at 3 h (9-fold) and then decreased over
time (4.5-fold at 8 h and 2.5-fold at 32 h, data not shown).
Addition of the protein synthesis inhibitor cycloheximide (20 µg/ml)
completely abolished TGF
-induced increases in SM
-actin mRNA
levels, suggesting that effects were dependent on new proteins of
ongoing synthesis of proteins with relatively short half-lives. TGF
also slightly induced nonmuscle (NM)
-actin mRNA expression.
The First 125 bp of the SM
To determine whether increases in SM
-actin gene transcription contributed to TGF
-induced increases in
SM
-actin mRNA expression, transient transfection studies were
performed in cultured rat aortic SMC using a construct containing 2.8 kb of the 5
-flanking sequence of the SM
-actin gene linked to a
promoterless CAT reporter gene (pProm/CAT) (29). Results demonstrated
that TGF
induced an ~7-fold increase in activity of pProm/CAT
above vehicle-treated controls (Fig. 2) suggesting that
TGF-induced increases in SM
-actin mRNA were due, at least in
part, to increased SM
-actin gene transcription. To identify the
minimal sequences within the pProm/CAT construct that were required for
TGF
responsiveness, we tested a series of deletion mutant
constructs. Transfection data indicated that the first 125 bp of the SM
-actin promoter were sufficient to confer full TGF
responsiveness. Previous analysis of the promoter regions between
125
bp and
2.8 kb demonstrated that it contained negative regulatory
elements required for cell-specific expression of the SM
-actin gene
(29). Inclusion of these upstream sequences, however, did not alter
TGF
-induced stimulation of SM
-actin gene expression in SMC.
Identification of a TGF
To identify potential TGF
response elements within the first 125 bp, we first performed a
sequence analysis to determine whether the initial 125 bp of the SM
-actin promoter contained any known TGF
response elements
(46-50). A 10-bp sequence spanning from
48 to
57 shared sequence
similarities with a previously reported TGF
response element
designated as TGF
inhibitory element (TIE) paradoxically found in
promoters of several genes whose expression were inhibited
by TGF
(51, 52; Table I). We analyzed this region by
site-directed mutagenesis to test its importance for TGF
responsiveness (Fig. 3A). The design of
mutants was based on those shown by Pietenpol et al. (51) to
abolish TGF
-mediated inhibition of the c-myc gene.
Results of transient transfection studies demonstrated that
TGF
-induced stimulation of the SM
-actin gene was nearly
completely abolished in all three mutant constructs (Fig.
3B). Mutations of this TGF
responsive control element
(henceforth designated "TCE") also reduced basal transcriptional
activity of p125 CAT by approximately half, even in the absence of
exogenous TGF
. We (53) and others (54) have previously shown that
cultured rat SMC produce active TGF
. Thus, it is possible that the
effect of mutation of the TCE on "basal" transcription was also
TGF
-dependent. To test this possibility, we incubated
SMC with a TGF
neutralizing antibody (raised in chicken, R&D) during
the entire period following transfection. Control cells were treated
with chicken IgG. In an attempt to maximize inhibition of autocrine
produced TGF
, high concentrations of TGF
antibodies (2 µg/ml)
were administered. Results demonstrated that addition of TGF
antibodies reduced the basal transcriptional activity of wild-type p125
CAT by ~50% as compared with controls (Fig. 4). In
contrast, TGF
neutralizing antibodies had no effect on the activity
of TCE mutants, although in the presence of neutralizing TGF
antibodies, basal transcriptional activity of the TCE mutants was about
50% lower than wild-type p125 CAT. Results of these studies indicate
that endogenously produced TGF
contributes to basal transcriptional
activity of p125 CAT. It cannot be ascertained from these data,
however, whether there is also a TGF
independent mechanism that
contributes to the reduced basal transcription seen with the TCE
mutants (Fig. 4), since it cannot be determined if complete
neutralization of TGF
was achieved over the entire time course of
the experiment.
|
Two Highly Conserved CArG Elements (A and B) Contained Within the First 125 bp of the SM
CArG
elements have been shown to be important for basal SM -actin
transcription in SMC (29) as well as for basal and TGF
-induced stimulation of the skeletal
-actin gene in cardiac muscle cells (30). Thus, we tested whether the CArG elements in the SM
-actin promoter were required for TGF
responsiveness. Mutation of either CArG A or CArG B, alone or in combination, completely abolished TGF
-induced increases in CAT activity in transient transfection assays (Fig. 5). Mutation of the CArG elements also
completely abolished TGF
responsiveness within the context of the
2.8-kb promoter construct, pProm/CAT (data not shown).
TGF
To characterize protein interactions with the TCE, EMSA were
performed with a 19-bp probe containing the TCE (henceforth designated as "TCE probe") and nuclear extracts from TGF- or
vehicle-treated SMC. Results of EMSA with the TCE used as a probe
showed a single shift band with nuclear extracts from TGF
-treated
SMC (henceforth designated as shift band 1, Fig. 6,
lane 2) that was barely detectable in extracts from
vehicle-treated cells (Fig. 6, lane 1). Addition of cold
double-stranded wild-type TCE competitor oligonucleotides inhibited
formation of shift band 1 (lanes 3 and 4),
indicating that shift band 1 represented a sequence-specific
protein-DNA complex. Competition with mutant TCEm2 cold double-stranded
competitor oligonucleotides did not affect formation of shift band 1 (lanes 5 and 6). No shift band was formed when
the mutant TCEm2 oligonucleotide was used as labeled probe (lane
13). These results are consistent with data obtained from
transfection studies showing that mutations of the TCE element resulted
in loss of TGF
responsiveness. Previous studies have shown binding
of the transcription factor AP-1 to the TIE in the transin/stromelysin
promoter (52). To determine if AP-1 binding contributed to formation of
shift band 1, cold oligonucleotides containing a canonical AP-1
consensus sequence were used in competition assays. AP-1
oligonucleotides did not affect formation of shift band 1 (lanes
7 and 8). Further evidence that AP-1 is not likely to
be responsible for formation of shift band 1 is provided by a study by
Angel et al. (55) which demonstrated that a T at position 1 of the AP-1 consensus sequence (TGAGTCAG) is critical for AP-1 binding.
Moreover, the TCE in the SM
-actin promoter shares only 4 bp (GAGT)
with the AP-1 consensus sequence and the T at position 1 is replaced by
a C. Taken together, these results suggest that AP-1 was not part of
shift band 1 (lanes 7 and 8).
The Nuclear Factor That Bound to the TCE Exhibited Binding Properties That Distinguished It from Sp1 or Sp3
Since part of
the TCE sequence (GGGAGGG) shares similarities with a transcription
factor Sp1 recognition site, and in addition Sp1 has been implicated in
TGF-mediated gene expression (47, 49, 50), additional experiments
were performed to determine whether Sp1 or an Sp1-like factor bound to
the TCE. Formation of shift band 1 was decreased by the addition of
high excesses of cold oligonucleotides containing a canonical
Sp1-binding site indicating that the nuclear protein responsible for
formation of shift band 1 could bind to a Sp1-binding site, albeit
weakly (Fig. 6, lanes 9-11). Several lines of evidence,
however, suggested that the factor binding to the TCE sequence was not
Sp1. First, incubation of labeled TCE probe with nuclear extracts from
TGF
-treated SMC or with large amounts of recombinant Sp1 resulted in
formation of shift bands (Fig. 7, compare lanes 6 and 9) with different mobilities. Second, the shift
complex formed by labeled TCE and recombinant Sp1 co-migrated with
those formed by labeled Sp1 probe and nuclear extracts from
TGF
-treated SMC (Fig. 7, lane 2). Both of these complexes
were supershifted by Sp1 antibodies (Fig. 7, lanes 3 and
10) indicating that the Sp1 antibodies specifically recognized human recombinant Sp1 and Sp1 present in nuclear extracts from rat aortic SMC. However, Sp1 antibodies did not affect shift band
1 formation, when labeled TCE probe was incubated with nuclear extracts
from TGF
-treated cells (Fig. 7, lane 7). These results strongly suggest that Sp1 was not the TGF
-induced nuclear factor binding to the TCE probe.
Recently, other members of the Sp1 transcription factor family have been cloned and sequenced (56). One of these family members, Sp3, has also been partially characterized and shown to bind to similar sequences as Sp1 (56). We therefore tested whether Sp3 might be responsible for formation of shift band 1. Sp3 antibodies did not affect formation of shift band 1, indicating that the TCE binding factor was not Sp3 (Fig. 7, lane 8). The ability of the Sp3 antibody to form supershifted complexes was verified using an Sp3 containing nuclear extract (data not shown).
TGFSince transfection
data provided evidence for functional importance of the CArG elements
for TGF inducibility of the SM
-actin gene, we tested whether
TGF
treatment affected binding to the SM
-actin CArGs. EMSA were
performed with labeled 20-bp CArG A or B oligos and nuclear extracts
from either TGF
- or vehicle-treated SMC. TGF
treatment markedly
enhanced binding to both CArG elements (Fig. 8,
lanes 2 and 5) when compared with vehicle-treated
controls. Consistent with our previous studies (29), two shift bands
were identified that supershifted with addition of SRF antibodies (data not shown). These results demonstrate that SRF is present in the protein-DNA binding complex formed with nuclear extracts from TGF
-treated SMC. TGF
-enhanced SRF binding was not associated with
changes in mobility as compared with the shift bands formed with
recombinant SRF, suggesting that TGF
did not induce formation of a
stable higher order complex, at least under the gel-shift assay
conditions employed in these studies (Fig. 8A, lanes 3 and 6). TGF
treatment also markedly enhanced binding activity
to the Wt 95 probe (Fig. 8B, lane 2) compared with vehicle
controls (lane 1). Addition of SRF antibodies (Fig.
8B, lane 3) led to formation of supershifted complexes and
the disappearance of bands 1 and 2, indicating that these two bands
contained SRF. The SRF antibodies employed also appeared to inhibit SRF
binding since bands 1 and 2 are not quantitatively supershifted. To
test whether TGF
-induced increases in SRF protein levels might have
contributed to the enhanced binding activity noted for the SM
-actin
CArGs, we performed Western analysis of cell lysates obtained from
TGF
-treated SMC. Results showed increased immunoreactive SRF in
TGF
-treated cells as compared with vehicle-treated cells (Fig.
9), suggesting that enhanced CArG binding following
TGF
treatment was due, at least in part, to increased SRF
expression.
TGF
The promoters of the SM MHC (rat, rabbit) and
h1 calponin (mouse) have recently been cloned and partially
sequenced (57-59). A computer-assisted search revealed that both
promoters contain TCE-like sequence motifs (Table I). Thus, we tested
whether TGF treatment of SMC also stimulated expression of these
genes. A comparative Northern analysis of control versus
TGF
-treated SMC showed that TGF
markedly increased both SM MHC
(Fig. 10A) and h1 calponin (Fig.
10B) mRNA levels. These results suggest that TGF
might act as a positive SMC differentiation factor by coordinately up-regulating expression of several SM differentiation marker genes.
The goal of the present study was to investigate the molecular
mechanisms whereby TGF up-regulates SM
-actin expression. Through
both deletion and mutational analysis, we identified the promoter
sequences required for TGF
activated SM
-actin transcription. Results showed that the first 125 bp of the SM
-actin promoter were
sufficient to confer TGF
-induced activation and that three positive-acting cis-elements contained within this region
were essential for this response, CArG box A at
62 and B at
112 and a TCE at
42. The fact that mutation of any one of these elements completely abolished transcriptional activity suggests that these elements operate interactively rather than independently to confer TGF
responsiveness.
Mutation of the TCE resulted in both loss of TGF induction of SM
-actin expression, and loss of binding of an as yet unidentified TGF
-inducible factor to the TCE, suggests that the TCE functions as
a positive-acting cis-element. Paradoxically, this SM
-actin TCE shares sequence similarities with a previously described
TIE found in a number of genes inhibited by TGF
(60), including transin/stromelysin (52) and c-myc (51). Of interest,
however, Pietenpol et al. (51) presented evidence suggesting
that TGF
-induced inhibition of c-myc transcription in
keratinocytes was mediated through down-regulation of a positive
trans-acting factor (51). As such, the TCE in the
c-myc promoter might also function as a positive-acting
cis-element as we observed for the SM
-actin TCE.
However, whether the TCE binding factors in SMC and keratinocytes are
related, and whether they are regulated in a cell-type specific manner,
will require further studies.
TGF responsiveness of the SM
-actin gene was also dependent on
two highly conserved CArG elements, A and B, that bind SRF (29).
Evidence for functional importance of CArG elements in TGF
-mediated
gene transcription has been shown previously in studies by MacLellan
et al. (30) who showed that the most proximal CArG box in
the skeletal
-actin promoter was essential for TGF
responsiveness
in cardiac myocytes whereas mutation of the more distal CArG elements
only partially inhibited TGF
responsiveness. Several lines of
evidence indicate that these genes are regulated differently by TGF
in cardiac myocytes versus SMC. First, TGF
-mediated activation of the skeletal
-actin gene was found to be dependent on
an M-CAT site which binds members of the TEF-1 family (61) in that
mutation of this site abolished basal and TGF
-inducible expression.
The M-CAT sites in the SM
-actin promoter, however, are located
upstream (
178 and
314) of the minimal sequence required for full
TGF
responsiveness. Second, the region on the skeletal
-actin
promoter shown to be TGF
-responsive does not contain a TCE or
TCE-like element. Third, there is evidence that
CArG-dependent regulation of these two genes is different
with respect to TGF
activation. For example, our results
demonstrated that TGF
markedly enhanced SRF binding to the CArG
boxes, and increased SRF protein expression. In contrast, MacLellan
et al. (30) did not observe TGF
-induced changes in SRF
binding to the skeletal
-actin SRE by gel shift analysis. It should
be noted, however, that the proximal SRE of the skeletal
-actin
promoter contains overlapping sites for SRE and YY1 and thus, SRF
binding to the proximal SRE could have been influenced by YY1 competing
with SRF binding.
Our observation that TGF increased SRF in SMC suggests that
TGF
-mediated activation of SM
-actin might be due in part to increased SRF expression and binding. Consistent with this, there is
extensive evidence demonstrating that increased SRF expression and
binding to CArG elements is associated with increased transcriptional activity of several genes including skeletal
-actin (62, 63), SM
-actin (64), and c-fos (65). The fact that TGF
-induced stimulation of SM
-actin mRNA expression was dependent on
ongoing protein synthesis (Fig. 1) is consistent with a role for
increased synthesis of SRF in mediation of TGF
inducibility of the
SM
-actin gene. Alternative mechanisms, however, must also be
considered whereby TGF
might mediate increased SRF binding. These
include post-translational modifications of SRF (66), and interaction with homeodomain containing transcription factors that have been shown
to modify binding of SRF to CArG elements (67). Of interest, we have
recently demonstrated that MHox, the murine homologue of the
homeodomain containing protein Phox 1 can potentiate binding of SRF to
CArG B (68). It remains to be determined, however, whether TGF
influences MHox expression and/or activity, and whether MHox is
involved in mediating TGF
effects on SM
-actin expression.
Our observations that TGF also stimulated increased expression of SM
MHC and h1 calponin raise the interesting possibility that
it may function as a positive differentiation factor for SMC. Moreover,
the fact that the promoters of each of these three genes contain paired
CArG elements and at least one TCE-like element (Table I) suggest a
potential common mechanism that might contribute to coordinate
regulation of expression of these genes during SMC differentiation.
Consistent with this, we have recently demonstrated functionality of
the two SM MHC CArGs at
1103 and
1221 by site-directed mutagenesis
(69). Moreover, with the possible exception of h1 calponin
(70) CArG elements are also required for maximal expression of a number
of other SMC differentiation marker genes including the SM-22
and
h-caldesmon gene in cultured SMC (71, 72). The effects of mutations of
the TCE-like elements on TGF
inducibility of the SM MHC,
h1 calponin, and other SMC differentiation marker genes
will need to be tested in further studies. Further evidence that TGF
might act as a positive differentiation factor in SMC is provided by
studies of Shah et al. (17), who demonstrated that TGF
treatment stimulated differentiation of pluripotent neural crest stem
cells into smooth muscle cells or SMC-like cells based on morphology
and induction of expression of SM
-actin and calponin. Moreover,
Grainger et al. (73) demonstrated that TGF
treatment
inhibited down-regulation of SM MHC isoforms in primary cultures of SMC
suggesting that TGF
might be involved in maintaining of SMC in a
more differentiated state. However, in contrast to our results, TGF
failed to induce re-expression of SM MHC in subcultured SMC. The reason
for these differences is not clear but may relate to culture
methodologies (14). In particular, the studies reported by Grainger
et al. (73) were done in the presence of serum, whereas ours
involved initial growth arrest in a defined serum-free medium prior to
TGF
stimulation.
A number of caveats need to be considered before concluding that TGF
might be a positive differentiation factor for SMC. First, our
observation that TGF
also stimulated small increases in NM
-actin
expression would appear to be inconsistent with TGF
being a positive
differentiation factor for SMC. However, in separate studies, we have
shown that TGF
specifically increased SM
-actin mRNA
expression when administered at lower concentrations than employed in
the present studies (i.e. 0.1 ng/ml versus 2.5 ng/ml).3 Second, one must consider the
possibility that the TGF
effects on SM differentiation marker
expression may simply be secondary to growth effects. However, this
seems to be highly unlikely, since TGF
had no effect on cell growth
under the experimental conditions used in the present studies which
involved initial growth arrest of SMC in a defined serum-free
medium.
In summary, the present study showed that the two CArG elements A and B
and a novel TCE were required for TGF responsiveness of the SM
-actin gene. Regulation appears to involve SRF binding to the CArGs
as well as an yet unidentified factor or factors that bind to the TCE
in a TGF
-dependent manner. Identification of this factor
may help to elucidate mechanisms that control SMC differentiation
during vasculogenesis and in diseases such as atherosclerosis and
injury-induced restenosis which are characterized by altered expression
of multiple SMC differentiation proteins as well as TGF
1.
We gratefully acknowledge the expert technical assistance of Diane Raines and Andrea Tanner.