Osteopontin Transcription in Aortic Vascular Smooth Muscle Cells
Is Controlled by Glucose-regulated Upstream Stimulatory Factor and
Activator Protein-1 Activities*
Miri
Bidder
§,
Jian-Su
Shao,
Nichole
Charlton-Kachigian,
Arleen P.
Loewy,
Clay F.
Semenkovich§, and
Dwight A.
Towler§¶
From the Divisions of Bone and Mineral Diseases and
§ Endocrinology, Diabetes, and Metabolism, Department of
Internal Medicine, Washington University School of Medicine, St. Louis,
Missouri 63110
Received for publication, June 22, 2002, and in revised form, August 12, 2002
 |
ABSTRACT |
The expression of the matrix cytokine osteopontin
(OPN) is up-regulated in aortic vascular smooth muscle cells (VSMCs) by diabetes. OPN expression in cultured VSMCs is reciprocally
regulated by glucose and 2-deoxyglucose (2-DG; inhibitor of cellular
glucose metabolism). Systematic analyses of OPN
promoter-luciferase reporter constructs identify a CCTCATGAC
motif at nucleotides 
80 to 72 relative to the initiation site that
supports OPN transcription in VSMCs. The region 83 to
45 encompassing this motif confers basal and glucose- and
2-DG-dependent transcription on an unresponsive promoter.
Competition and gel mobility supershift assays identify upstream
stimulatory factor (USF; USF1:USF2) and activator protein-1 (AP1;
c-Fos:c-Jun) in complexes binding the composite CCTCATGAC element.
Glucose up-regulates both AP1 and USF binding activities 2-fold in A7r5
cells and selectively up-regulates USF1 protein levels. By contrast,
USF (but not AP1) binding activity is suppressed by 2-DG and restored
by glucose treatment. Expression of either USF or AP1 activates the
proximal OPN promoter in A7r5 VSMCs in part via the
CCTCATGAC element. Moreover, glucose stimulates the transactivation
functions of c-Fos and USF1, but not c-Jun, in one-hybrid assays.
Mannitol does not regulate binding, transactivation functions, USF1
protein accumulation, or OPN transcription. Thus, OPN gene transcription is regulated by USF and AP1 in aortic VSMCs, entrained to changes in cellular glucose metabolism.
| |
INTRODUCTION |
Individuals with diabetes are characteristically afflicted with
medial artery calcification (1-4). The pathobiology of
diabetes-associated calcific vasculopathy is poorly understood. Medial
calcification occurs via a mineralization process that does not form
cartilage, with calcium deposition occurring in the collagenous
extracellular matrix associated with matrix vesicles adjacent to
arterial VSMCs1 (5).
Initially thought to be benign, medial calcification has emerged as one
predictor of cardiovascular mortality (2, 3). The mechanism whereby
medial calcification conveys excess mortality are unknown, but
vascular stiffening prevents the compensatory vasodilatation that
decreases afterload and myocardial oxygen consumption with exertion,
thus exacerbating myocardial ischemia from concomitant atherosclerotic
coronary disease (6-10).
We recently reported the development and initial characterization of a
murine model of diet-induced, insulin-resistant diabetic calcific
vasculopathy (11). Feeding LDLR / mice high fat diets induces
insulin-resistant diabetes and dyslipidemia, with associated mineral
deposition in the fibrosa of the valve leaflets and the arterial wall.
As initially proposed by Demer and colleagues (12, 13) from studies of
atherosclerotic plaques, our data suggest that an active osteogenic
program is initiated during aortic calcification by the dysmetabolic
stimuli of diabetes that promote medial calcific vasculopathy (11).
Aortic expression of the bone matrix protein osteopontin (OPN) is
up-regulated and is detected in several aortic cell types including
activated macrophages of atheroma, vascular smooth muscle cells of the
tunica media, and adventitial and valvular fibrosal cells (11, 14).
OPN is a multifunctional protein produced by osteoblasts during
skeletal development and in VSMCs, monocytes, and T-cells in response
to biological stressors (14, 15). As an extracellular matrix protein,
OPN controls cell migration mediated by
v
3 integrin receptors (14, 16). Via
signaling functions dependent upon phosphorylation, OPN inhibits
mineral deposition (17). OPN also functions as a proinflammatory
cytokine produced by stimulated monocytes and T-cells (14), augmenting
monocyte/macrophage activation via up-regulation of interleukin-12 and
suppression of interleukin-10 (18). Very recently, Mori and
co-workers (19) extended our work (11) to analyses of human diabetic
calcific vasculopathy. They identified that immunoreactive OPN was in
fact up-regulated in medial vascular smooth muscles cells in arteries
of diabetic patients (19). Highly relevant to arterial vasculopathy of
diabetes (20), Thompson and co-workers (21) have demonstrated that the
adventitial mesenchymal cell population is migratory, moving into the
tunica media and neointima in response to biomechanical injury. A role
is thus proposed for OPN in the migration of VSMCs and adventitial
mesenchymal progenitors in the diseased aorta. Therefore,
the enhanced expression of OPN in the arterial vasculature in response
to diabetes is likely to participate in disease progression, regulating
macrophage functions, adventitial cell migration, and calcium
deposition (17).
The goal of this study was to characterize transcription factor
complexes that confer basal and regulated OPN gene
expression in aortic VSMCs relevant to the pathobiology of diabetic
calcific vasculopathy. We have identified that OPN gene
expression is differentially regulated by glucose and 2-deoxyglucose
(2-DG) in cultured primary aortic mesenchymal cells that express VSMC
-actin. We cloned the 2-kb mouse OPN promoter (1976 to
+78, numbered relative to the start site of transcription) (22) and
evaluated the activity in A7r5 rat aortic VSMCs, a rodent cell line
that faithfully reproduces the gene expression protein elicited by
aortic VSMCs in vivo (23, 24). Glucose (
5 mM)
selectively up-regulates activity of the OPN promoter. By
contrast, mannitol, an osmotic control, and 3-O-methyl glucose, a nonmetabolized glucose an analog, have no effect. We identified that nucleotides 83 to 46 relative to the transcription initiation site encompass information necessary and sufficient for high
level transcriptional activity in VSMCs. This 38-bp fragment of the
OPN promoter confers basal and glucose
metabolism-dependent transcription onto an unresponsive
heterologous minimal promoter. A CCTCATGAC motif at 80 to 72 that
is recognized by USF1, USF2, c-Fos, and c-Jun supports OPN
promoter activity in VSMCs. These complexes are regulated by glucose.
Both USF1 binding and protein accumulation are reciprocally regulated
during the manipulation of cellular glucose metabolism with glucose and
2-DG. Co-transfection studies demonstrate that USF (USF1:USF2) and AP1
(cFos:cJun) support OPN transcription and that the
transactivation functions of c-Fos and USF1 are specifically enhanced
by glucose. Thus, basal and regulated protein-DNA interactions at the
CCTCATGAC motif at 80 to 72 participate in OPN gene
expression responses in VSMC during glucose-dependent
initiation of diabetic vasculopathy. USF and AP1 contribute to the
glucose-dependent regulation of arterial VSMC OPN
expression in diabetes.
| |
EXPERIMENTAL PROCEDURES |
Reagents, Antibodies, and Cell Culture--
Tissue culture
supplies and custom synthetic oligodeoxynucleotides were obtained from
Invitrogen. Reagents for protein preparation were obtained from
Pierce, Sigma, and Fisher. Radionucleotides were obtained from
Amersham Biosciences. A7r5 cells (ATCC CRL-1444) were maintained as
described previously (25, 26). Primary aortic adventitial mesenchymal
cells were obtained from aortic explants essentially as detailed
previously (27) but using C57Bl/6 mice in lieu of rats. Cultures were
passaged in DMEM with 4.5 g/liter glucose, 4 mM glutamine,
and 10% fetal calf serum (and penicillin-streptomycin) to maintain the
phenotype (25). About 16 confluent 10-cm-diameter tissue culture dishes
were obtained/eight 2-cm segments of murine aortae. Antibodies were
purchased from Santa Cruz Biotechnology; specific antibodies include
USF1 (sc-8983 or sc-229), USF2 (sc-861), c-Fos (sc-253 or
sc-52), FosB (sc-7203), Fra1 (sc-183), Fra2 (sc-171), c-Jun (sc-44 or
sc-45), phospho-c-Jun (sc-822), JunD (sc-74), Smad 1/5 (sc-6031), Oct1
(sc-232), Oct2 (sc-233), Oct4 (sc-9081), Gli1 (sc-6152 or sc-6153), and
CTCF (sc-5916) or BTEB2 (sc-12998).
Cell Culture under Conditions of Varying Glucose and
2-Deoxyglucose Concentrations--
A7r5 aortic VSMCs were maintained
with DMEM containing 1 g/liter (5.5 mM) glucose. Treatment
of A7r5 cells with either glucose or mannitol was performed in
glucose-free DMEM, 10% glucose-free FCS (dialyzed against glucose-free
DMEM) supplemented with either sterile water (0.1% v/v, vehicle; 0 mM glucose) or carbohydrates (5 mM glucose or
mannitol, 50 mM glucose or mannitol, 100 mM
glucose or mannitol as indicated). The nonmetabolized derivative of
glucose, 3-O-methylglucose (3-OMG) was added as a control.
Unlike A7r5 cells, primary aortic VSMCs do not adapt well to ex
vivo culture under completely glucose-free conditions even with 4 mM glutamine provided as an alternative carbon source (25,
28).2 Therefore, to permit a
direct comparison of OPN mRNA accumulation between A7r5 and primary
aortic cell cultures, cohorts were cultured in the presence of 2-DG, a
competitive inhibitor of cellular glucose uptake and
intracellular glucose metabolism (29), in the presence of
glucose-free DMEM with undialyzed FCS (final glucose concentration ~0.5 mM glucose) to pharmacologically induce acute and
reproducible extremes in cellular glucose tone. Experiments designed to
examine the reversibility of 2-DG actions by glucose were carried out in the presence or absence of 2 mM 2-DG.
Real-time Fluorescence RT-PCR Measurements of OPN and Msx2 Gene
Expression and Northern Blot Analyses--
Primary aortic
mesenchymal cells obtained from wild type adult C57Bl/6 mice were
passaged once and then plated at 80% confluence in 10-cm tissue
culture dishes. Subsequently, cells were treated for 24 h in 10%
undialyzed FCS with glucose-free DMEM (~0.5 mM residual
glucose from the FCS) or in media supplemented with 2 mM
2-deoxyglucose, 4 mM 2-deoxyglucose, 30 mM
glucose, or 30 mM glucose + 2 mM
2-deoxyglucose. At the end of the treatment period, total RNA was
extracted as described previously (30). An Applied Biosystems GeneAmp
5700 sequence detection system using Sybr Green dye binding to PCR
product was used to quantify osteogenic mRNA accumulation via
fluorescence RT-PCR (31). Primer Express 1.0 software was used to
design amplimers from murine OPN and Msx2 cDNA sequences. Amplimers for OPN are: 5'-GTA TTG CTT TTG CCT GTT
TGG-3' and 5'-TGA GCT GCC AGA ATC AGT CAC T-3'. Amplimers for murine
Msx2 are: 5'-TCC CAG CTT CTA GCC TTG GA-3' and 5'-CAG CCC
GCT CTG CTAT GG-3'. Commercially available primers from Applied Biosystems were used to quantify GAPD expression for
normalization in RT-PCR (TaqMan Rodent GAPDH Control Reagent, PE
Biosytems, Palo Alto, CA). Standard curves were generated using
OPN and Msx2 DNA standards. Data are presented as
the mean ± S.D. results from three independent replications of
two independent experiments. Total cellular RNA was isolated from VSMCs
and Northern blot analysis carried out as described previously (30,
32).
Immunohistochemical Staining of Cultured Cells--
Cultures of
mouse aortic adventitial cells and A7r5 rat aortic VSMCs were evaluated
for expression of OPN and VSMC -actin by immunohistochemistry using
the streptavidin-biotin-immunoperoxidase method (33). A mouse
monoclonal antibody to human VMSC -actin was obtained from a
commercial source (product code CBL 171, clone asm-1, Cymbus
Biotechnology). The osteopontin antibody utilized was MPIIIB10 (1),
which has been used previously to localize vascular OPN expression in
calcified aortic specimens (34). This mouse monoclonal was developed by
Michael Solursh and Ahnders Franzen (Dept. of Biological Sciences,
University of Iowa, Iowa City) and was obtained from the Developmental
Studies Hybridoma Bank developed under the auspices of the NICHD,
National Institutes of Health, and maintained by the Department of
Biological Sciences, University of Iowa. Immunostaining of cultured
cells was performed using the DAKO Animal Research Kit with peroxidase
histochemistry (DAKO Corp., Carpinteria, CA) as per the manufacturer's
protocol, using diaminobenzidine with H2O2 as
the chromagen substrate. Microscopic images of immunohistochemically
stained cells were captured with a Nikon Coolpix 990 3.3 megapixel
digital camera mounted to a Nikon Eclipse TS100 inverted microscope
(Melville, NY).
Eukaryotic Expression Constructs, OPN Promoter-Reporter
Constructs, and Transient Transfection Assays--
Mouse USF1 cDNA
was obtained by PCR amplification (which also introduced convenient 5'-
and 3' linkers) from commercially available murine heart cDNA
(Clontech Marathon-Ready cDNA;
Clontech, Palo Alto, CA) and subcloned into the
KpnI/BamHI sites of pcDNA3 (Invitrogen). In
addition, coupled in vitro transcription/translation (Promega, Madison, WI) was used to verify protein production using techniques detailed previously (35). The same techniques were used to
develop the eukaryotic expression constructs for c-Fos and c-Jun in
pcDNA3. The expression vector for USF2 (kindly provided by Dr. M. Sawadogo) has been described previously (36). PATHDETECT vectors for
eukaryotic expression of Gal4DBD fusion proteins (pFACMV) and the
Gal4RE-LUC reporter (pFRLUC) were purchased from Stratagene. pFACMV was
used to assemble the eukaryotic expression plasmid for
Gal4DBD-USF1. The synthesis of 1976 OPNLUC (mouse OPN promoter fragment 1976 to +78 in pGL2 Basic; numbering as per Yamamoto and
colleagues (22), GenbankTM accession no. D14816) was
obtained by PCR using mouse genomic DNA as a template, applying
techniques described previously (37). OPN promoter
deletions, point mutants, and heterologous promoter constructs were
generated using methods as detailed (38). RSVLUC (38) and CMVLUC (39)
promoter-reporter constructs have been described previously. All
constructs were sequenced to verify fidelity (Applied Biosystems Prism
Dye Terminator Kit, Foster City, CA). A7r5 aortic VSMCs were
transfected using LipofectAMINE (Invitrogen). CMV -galactosidase
(300 ng) was included as an internal control for transfection
efficiency. One day after transfection, cultures were rinsed with
glucose-free DMEM and then fed with glucose-free DMEM, 10% dialyzed
FCS serum supplemented with glucose, mannitol, 3-OMG, or 2-DG as
indicated. After 3 days, cellular luciferase and -galactosidase
activities were measured as detailed previously (38, 40). Statistical
analyses were carried out as previously described (39).
Electrophoretic Mobility Gel Shift Assays, Supershift Assays,
Immunoprecipitation, and Western Blot Analyses--
Crude extracts for
gel shift analyses were prepared from control, glucose, and/or 2-DG
treated cells (see above) as detailed previously (26). The radiolabeled
OPN promoter duplex oligo fragments (see Fig. 4) were used
to detect DNA-protein interactions by gel shift and supershift as
described. Western blots were carried out with chemiluminescent
immunodetection (Tropix, Bedford, MA) as described previously (26,
41).
| |
RESULTS |
OPN Gene Expression in Cultured Aortic VSMC Cells Responds to
Glucose and to Pharmacological Manipulation of Cellular Glucose
Metabolism--
Previously, we identified that OPN gene expression is
up-regulated in aortic VSMCs, adventitial cells, and macrophages of LDLR/ mice fed diabetogenic diets. In situ hybridization
studies revealed that aortic mesenchymal OPN expression
overlaps the pattern of VSMC -actin in adventitial cells, mural
VSMCs, and valvular fibrosal cells (11). We wished to establish a cell
culture model for studying OPN transcription in aortic
mesenchymal cells. Therefore, we studied the expression and regulation
of OPN in primary aortic mesenchymal cell cultures and the
A7r5 aortic VSMC line, a rat aortic VSMC line that faithfully
recapitulates the transcriptional regulatory features of arterial
smooth muscle cells (23, 24). As in A7r5 VSMCs (Fig.
1A),
immunohistochemical analysis of primary aortic mesenchymal cells (Fig.
1D) demonstrates robust staining of >80% of cells with
VSMC-specific -actin, indicative of smooth muscle phenotype (42,
43). Immunohistochemistry further identified the expression of
OPN in both A7r5 cells and primary murine aortic cells; in A7r5 cells,
the prominent perinuclear Golgi staining described by others (33) was
readily apparent. Northern blot and RT-PCR analyses confirmed
expression of OPN mRNA in these cells and identified
response to alterations in glucose treatment and metabolism. As shown
in Fig. 2A, treatment of
primary cultures of aortic mesenchymal VSMCs up-regulated
OPN mRNA accumulation by ~2-fold. By contrast,
treatment with 2-DG dose dependently and selectively down-regulated
OPN mRNA accumulation by 80% (Fig. 2B;
GAPD normalized) as quantified by fluorescence RT-PCR
analyses. The Msx2 gene was regulated in culture in a
completely different manner; glucose suppressed Msx2
mRNA accumulation, and 2-DG actually augmented Msx2
expression 3-fold (Fig. 2B), demonstrating that the acute
exposure to 2-DG is not toxic to primary aortic cells. Although
expressed at much higher levels, OPN mRNA accumulation was similarly regulated by 2-DG and glucose in A7r5 cells (Fig. 2C; and see below), and 5-50 mM mannitol did
not regulate OPN in this VSMC line (data not shown, and see
below). Thus, as noted in vivo in the VSMCs and adventitial
cells of diseased aortae (11, 15), cultured aortic mesenchymal cells of
the VSMC lineage express the OPN gene under basal
conditions. The aortic mesenchymal cell OPN gene in
vitro is responsive to glucose concentrations and pharmacological
manipulation of cellular glucose metabolism (29).
View larger version (104K):
[in this window]
[in a new window]
|
Fig. 1.
Expression of immunoreactive VSMC
-actin and osteopontin in cultured aortic
mesenchymal cells. Cultures of A7r5 aortic VSMCs
(A-C) and primary murine aortic mesenchymal cells
(D-F) were fixed with methanol:acetone and immunostained
for VSMC -actin (A and D, -Act)
or OPN (B and E). Note that although cell
morphology differs, both A7r5 aortic VSMCs and primary aortic
mesenchymal cell cultures express VSMC -actin (A and D)
and OPN (B and E). C and
F, negative controls (CONT).
|
|
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 2.
Regulation of OPN gene expression
in aortic mesenchymal cells by manipulation of cellular glucose
homeostasis. A, Northern blot analyses of RNA from cultured
primary aortic VSMCs treated with 32 media supplemented
with vehicle (Control; 0 mM), 30 mM
glucose, or 2 mM 2-DG for 24 h as described under
"Experimental Procedures." Note that glucose treatment up-regulates
OPN mRNA accumulation 2-fold (GAPD
normalized), whereas 2-DG suppresses expression slightly. B,
primary murine aortic mesenchymal cells were treated with the indicated
concentrations of 2-DG or glucose either alone or in combination for
24 h in an independent experiment. OPN mRNA
accumulation was quantified by real time fluorescence RT-PCR
(normalized to the expression of GAPD mRNA). Note that
2-DG dose-dependently suppresses OPN mRNA
accumulation in primary murine aortic cells, whereas glucose increases
expression. By contrast, the gene for Msx2 is up-regulated
by 2-DG and suppressed by glucose in these same cell cultures.
B, A7r5 rat aortic VSMCs were cultured in glucose-free
medium, 10% FCS supplemented either with vehicle (water) or the
indicated concentrations of the solutes glucose and/or 2-DG. RNA was
prepared from A7r5 VSMCs treated with vehicle (lane 1), 2 mM 2-DG (lane 2), 4 mM 2-DG
(lane 3), 30 mM glucose (lane 4), or
2 mM 2-DG with 30 mM glucose (lane
5) and probed for OPN and GAPD
expression.
|
|
Basal OPN Promoter Activity in A7r5 Aortic VSMCs Is Dependent upon
an Intact CCTCATGAC Motif at Nucleotides 80 to 72 Relative to the
Transcription Initiation Site--
A7r5 cells are an
immortalized line derived from rat aortae that faithfully recapitulates
transcriptional regulation of aortic VSMCs in vivo. These
cells are readily grown and transfected, thus making A7r5 an excellent
cell culture system for detailed studies of promoter regulation in
aortic VSMCs (23). As noted, like primary aortic mesenchymal cell
cultures, this clonal line expresses VSMC -actin and OPN.
We wished to identify the promoter elements that regulate OPN gene
transcription in aortic VSMCs under basal conditions and in response to
the changes in glucose concentrations potentially relevant to diabetic
vasculopathy. Therefore, we cloned the 2-kb murine OPN
promoter fragment 1976 to +78 (numbering as per Yamamoto and
colleagues (22)) upstream of the LUC reporter gene. We then generated a
systematic series of 5'-deletion constructs, and evaluated promoter
activity after transient transfection of A7r5 cells by quantifying
cellular luciferase enzyme activity as described under "Experimental
Procedures." As shown in Fig. 3, basal
activity was dependent upon three regions: 636 to 135, a region
between the vitamin D-responsive enhancer at 764 to 747 (44) and
the proximal promoter; 107 to 83, encompassing an E-box and Sp1
regulatory region (45); and a new region between 83 to 61 relative
to the transcription initiation site. Notably, under these conditions,
the Runx and Ets cognates between 135 to 107 necessary for
supporting activity in osteoblasts (46) were not required for activity
in A7r5 aortic VSMCs. Basal activity of the proximal OPN
promoter was greatly dependent upon elements encompassed between
nucleotides 83 to 61. Although 83 OPNLUC (83 to +78) was highly
active, 61 OPNLUC (61 to +78) exhibited transcriptional activity only
slightly above background signal. Inspection of the promoter sequence
between 83 and 61 revealed an 11-bp region (CCTCATGACAC) encoding
three potential elements: a CCTC motif for recognition by the zinc
finger factors such as CTCF and the Kruppel family (47, 48); a CCTCATGA
motif resembling both atypical E-box (CCTNNTG) (49) and Oct (CTCATGA) cognates (50); and an overlapping TCATGAC representing a potential AP1
cognate (51) for recognition by Fos:Jun or ATF:Jun dimers (Fig.
4). To assess the contributions of each
element to basal activity, dinucleotide and trinucleotide thymidine
mutations were introduced into the wild type sequence (CCTCATGACACA) to
map the regions and elements contributing to basal promoter activity
(mutant 1, TTTCATGACAC; mutant 2, CCTCATTATAC; mutant 3, CTCATGATTT;
see summary in Fig. 4). These point mutations were introduced in the context of 83 OPNLUC (83 to +78) and compared with the activity of
the wild type promoter fragment (CCTCATGACACA). As shown in Fig.
3B, the introduction of thymidine mutations at
nucleotides 80 and 79 (MUT#1; Figs. 3B
and 4) and nucleotides 72 to 70 (MUT#3, Figs.
3B and 4) decreased activity by ~40%. Moreover, a
dithymidine transition at nucleotides 74 and 72 (MUT#2,
Fig. 3B) reduces basal promoter activity by ~70%. Thus,
OPN promoter activity in A7r5 aortic VSMCs is dependent upon
the intact CCTCATGACAC motif in the OPN promoter region 80 to
70.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 3.
Basal activity of the OPN
promoter in A7r5 aortic VSMCs is supported by promoter regions
636 to 135 and 107 to 61 dependent upon an intact CCTCATG motif
at nucleotides 80 to 75. A series of OPN
promoter-luciferase reporter constructs were transiently transfected
into A7r5 aortic VSMCs, and promoter activity was quantified by
assaying luciferase activity. CMV promoter-driven
-galactosidase was included to control for transfection efficiency
as outlined under "Experimental Procedures." A, data are
presented as the mean ± S.D. of OPN promoter
activity observed, with references to the basal activity of 107 OPNLUC
(OPN promoter 107 to +78). The promoter region 636 to 135,
lacking the vitamin D response element at 764 to 747, significantly
contributes to basal OPN promoter activity in A7r5 VSMCs.
Note, however, that in the proximal promoter 5'-deletion of 107 to
61 results in an ~15-fold reduction in basal activity with
approximately half of the basal activity conferred by the region 107
to 83 and half by the region 83 to 61. B, a series of
thymidine mutations was introduced into the proximal OPN
promoter fragment 83 to +78 and analyzed for transcriptional activity
after transient transfection of A7r5 VSMCs. Mutations were introduced
into the CCTCATGACACA motif at 80 to 69 (See Fig. 4) that mapped
elements supporting basal activity. Note that thymidine substitutions
in these region all diminished basal OPN promoter activity;
MUT #2 (Fig. 4) was most severe, decreasing basal activity by 70%
relative to the wild type promoter.
|
|
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 4.
Sequence of the proximal OPN
promoter region 85 to 49 and mutants analyzed. Putative
cognates for protein-DNA interactions in the OPN promoter in
this region are indicated. See text for details. WT, wild
type.
|
|
The Proximal OPN Region Is Differentially Regulated by Glucose and
2-Deoxyglucose in A7r5 Aortic VSMCs--
To determine whether the
aortic VSMC OPN gene expression responses observed with pharmacological
manipulation of cellular glucose uptake and metabolism reflect, in
part, changes in OPN promoter activity, we examined
the effects of 2-DG and glucose treatment on OPN
promoter-luciferase reporter activity in transiently transfected A7r5
VSMCs. As shown in Fig.
5A, glucose
treatment up-regulated 2-kb OPN promoter activity by
~3-fold. Induction is observed with both 5 and 50 mM
glucose. Induction was specific for glucose, because mannitol exposure
at identical concentrations, an osmotic control, elicits no activation.
Moreover, the transcriptional response to glucose was
promoter-specific, because the CMV (Fig. 5, A and
D) and RSV promoters (Fig. 5C; see below) were
not significantly responsive. Induction is dependent upon glucose
metabolism, because 3-OMG, a nonmetabolized glucose analog, does not
recapitulate or inhibit induction of 2-kb OPNLUC (Fig. 5B).
Moreover, competitive inhibition of intracellular glucose metabolism
with 2-DG decreases OPN promoter activity, and inhibition
was partially reversed with treatment with 60 mM glucose
(Fig. 5B). Again, the effects of short term treatment with
2-DG was promoter-specific, because the CMV promoter is not inhibited
(data not shown and Fig. 5D; see below) The effects
of the systematic series of 5'-truncated OPN promoter-LUC
constructs that we described above were tested for induction by glucose
in A7r5 cell to map this response. As occurs with 1976 OPNLUC, 636 OPNLUC (not shown), 135 OPNLUC, and 83 OPNLUC (83 to +78; Fig.
5C) were all induced by glucose treatment; 50 mM
glucose was used because this provided the maximal response observed
(Fig. 5A). By contrast, 61 OPNLUC, possessing the CCAAT box,
TATA box, and initiator region element of the OPN
gene (61 to +78), was unresponsive (Fig. 5B). To confirm
that the proximal region of the OPN promoter confers
transcriptional regulation in response to manipulation of cellular
glucose tone, we placed the 39-bp OPN fragment 83 to 45
upstream of the RSV minimal promoter. OPN-(83/45)
encompasses the CCTCATGACAC motif identified above as necessary for
robust basal activity in the critical region between 83 and 61 and
the CCAAT box at 53 to 49 (bottom strand). As shown in Fig.
5C, OPN-(83/45)RSVLUC activity was 20-fold greater than
that of the minimal RSVLUC in A7r5 VSMCs. Consistent with our deletion
analyses, the OPN promoter fragment 83 to 45 conferred glucose
responsiveness upon the unresponsive RSV minimal promoter (TATA box and
initiator region only; Fig. 5C). Moreover, as shown in Fig.
5D, the OPN promoter fragment 83 to 45
conferred 2-DG inhibition as well as glucose induction upon the
unresponsive RSV minimal promoter. Glucose treatment again stimulated
transcriptional activity ~2-fold; 2-DG suppressed activity by 60%,
and 50 mM glucose reversed 2-DG suppression (Fig.
5D). Again, activities of the RSV and CMV promoters were not
inhibited by 2-DG in A7r5 VSMCs (Fig. 5, C and
D). Thus, the OPN promoter region 83 to 45
assembles transcriptional complexes that support OPN
promoter activity in A7r5 aortic VSMCs; this region participates in
both basal transcriptional activity and regulation in response to
alterations in cellular glucose exposure and cellular metabolism.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 5.
OPN promoter elements encoded by
nucleotides 83 to 61 convey transcriptional responses to glucose
and 2-deoxyglucose. A, A7r5 aortic VSMCs were transiently
transfected with the 2-kb OPN promoter-luciferase reporter (1976 to
+78) as described in the legend to Fig. 3. Two days after
transfection, cells were treated for 24 h with the indicated
concentrations of glucose or mannitol, and promoter activity was
measured by assaying luciferase as described in the legend to Fig. 3.
OPN promoter activity is specifically up-regulated by glucose but not by mannitol (osmotic control). The
CMV (A) and RSV (C and
D) promoters are not significantly regulated. Induction is
dependent upon glucose metabolism; the nonmetabolized glucose analog
3-OMG is inactive (B), and 2-DG specifically inhibits
glucose-dependent activation. Deletion analyses map the
response to 83 to 61 (C) confirmed in heterologous
promoter assays. The OPN promoter region 83 to 45
confers basal activity responsive to manipulation of cellular glucose
homeostasis to the unresponsive RSV promoter (D).
See text for details.
|
|
VSMC Protein-DNA Interactions at the CCTCATGAC Motif--
To
initiate characterization of the DNA-protein interactions assembled by
the OPN proximal promoter, we performed electrophoretic mobility gel shift assays. Duplex synthetic oligodeoxynucleotides corresponding to OPN promoter fragment 85 to 65 were
radiolabeled with 32P and OPN promoter
DNA binding activity in A7r5 extracts from cells cultured as described
previously (26, 37). As shown in Fig. 6,
three closely migrating complexes (labeled A, B, and C) were resolved and visualized after autoradiography
(lane 1). (A much more rapidly migrating, variably
visualized complex of low specificity is also observed occasionally;
this is commonly seen with other oligos as well in this protocol (26)).
Cold homologous duplex oligo competed for formation of all three
complexes (Fig. 6, lanes 1-3). The cognate for AP1 binding
in the collagenase (MMP1) promoter (26) abrogated formation of complex
C (Fig. 6, lanes 4-6) and diminished complex A, indicating
the presence of AP1-like protein-DNA interactions in these two
complexes. However, complex B was unaffected by the AP1 oligo,
indicating a different binding specificity (Fig. 6, lanes
4-6). To derive the connections between the OPN promoter
sequences between 80 and 70 necessary for robust basal promoter
activity in A7r5 cells and in the assembly of specific A7r5 protein-DNA
interactions, duplex oligonucleotides containing mutations
corresponding to OPN promoter mutants MUT #1, MUT #2, and
MUT #3 (Fig. 4) were examined as cold competitors. A duplex oligo
possessing the CC
TT transition of MUT #1, which decreases
OPN activity by ~40% (Fig. 5B) and selectively
disrupts the putative E-box (Fig. 4; see below), can still compete for complexes A and C but not complex B (Fig. 6, lanes 7-9). By
contrast, a duplex oligo containing the MUT #2 alterations, which most
severely lesions basal OPN promoter activity (Figs.
3B and 4), did not compete for any of the three complexes
formed (Fig. 6, lanes 10-12; competition studies with cold
MUT #2 consistently augmented total binding activity, data not shown).
Like the wild type sequence, MUT #3 competed for formation of all
complexes (Fig. 6, lanes 13-15), indicating that at the
excess oligo concentrations used in competition assays, the intact
cognate CCTCATGAC was sufficient to compete for assembly of complexes.
Thus, protein-DNA interactions necessary for formation of complexes A,
B, and C on OPN promoter region 80 to 72 are necessary
to support robust basal OPN promoter activity in A7r5
myoblasts. Point mutations that perturbed the TCATGAC AP1-like
interactions alter recognition by complex C, a DNA-protein
interaction specifically disrupted by competition with an authentic AP1
cognate. Formation of complex A was also dependent upon sequences in
this element that overlapped the AP1 cognate, because nucleotide
substitutions that targeted this region precluded activity in
competition assays. Complex B DNA-protein interactions,
which overlap yet are distinct from the AP1-like interactions,
required an intact E-box-like motif; immunological probing confirmed
the presence of distinct protein-DNA complexes (see below). DNA-protein
interactions altered by MUT #2 simultaneously exert the most
significant deleterious effects on formation of specific DNA-protein
complexes (Fig. 6, lanes 10-12) and reductions in
OPN promoter activity (Fig. 3B).
View larger version (97K):
[in this window]
[in a new window]
|
Fig. 6.
The OPN promoter region 85 to
65 assembles three protein-DNA complexes. Extracts were prepared
from A7r5 cells grown under standard (25 mM glucose)
conditions, and DNA binding activities recognizing the
radiolabeled OPN promoter region 85 to 65 were detected
by gel shift assay as described under "Experimental Procedures."
DNA binding specificity was evaluated by competition with cold
homologous oligo (lanes 1-3), the interstitial collagenase
(MMP1) AP1 cognate (lanes 4-6), or three
OPN promoter mutants (lanes 7-9,
10-12, and 13-15) corresponding to the mutants
outlined in Fig. 4. Lanes 1, 4, 7, 10, and
13, no cold competitor; lanes 2, 5, 8, 11, and 14, indicated cold competitors at 10-fold molar
excess; lanes 3, 6, 9, 12, and 15,
indicated cold competitors at 30-fold molar excess. Three closely
migrating specific complexes are visualized in the gel shift
(lane 1). exposure times were shortened to facilitate
visualization of all three complexes. Complexes C and A (to a lesser
extent) both are reduced by competition with the AP1 cognate,
indicating AP1-like DNA-protein interactions in these two complexes.
Further note that complex B is not disrupted by this oligo (lanes
4-6) nor by OPN promoter fragment MUT #1 (lanes 7-9),
which contains mutations in the 5' portion of a putative E-box (Fig.
4). None of these complexes were inhibited by competition with the
OPN promoter fragment mutant 2 (lanes 10-12)
possessing thymidine substitutions that simultaneously destroys both
the E-box and AP1 cognates in this region. As observed for the wild
type oligo (lanes 1-3) mutant 3 was able to compete for
formation of all three complexes (lanes 13-15). See
"Results" for details.
|
|
Immunological Supershift Analyses Demonstrate the Presence of
USF and AP1 Factors in A7r5 DNA Binding Activities That
Recognize and Regulate the OPN CCTCATGAC Motif--
As noted above, in
addition to the TCATGAC AP1 motif, inspection of the OPN
promoter region 85 to 65 revealed an overlapping CCTCATG motif at
80 to 74 that resembles that of the CANNTG motif of classical E-box
cognates. The competition studies indicate that complex B specifically
recognizes the CCTCATG motif in the OPN promoter. Because
(a) the E-box binding factor USF1 has been shown to
recognize other distinct elements in the OPN promoter in
VSMCs (45), and (b) USF1 and USF2 have been shown to mediate hepatic glucose responses (52, 53), we immunologically probed the OPN
promoter binding complex B with anti-USF antibodies in gel supershift
assays as described under "Experimental Procedures." For these
experiments (Figs. 7), gel shifts were
carried out with radiolabeled OPN-(85/65) duplex oligo
and simultaneously in the presence of cold AP1 cognate to eliminate
band C; this permitted unobscured visualization of band B responses to
immunological probing. Because this region of the OPN
promoter also contains potential cognates (CCTC, CTCATGA) for
recognition by zinc finger and octamer factor, respectively, we
immunologically probed for these specific protein-DNA interactions as
well. As shown in Fig. 7, in this systematic screen, only the anti-USF1
antibody almost completely disrupted the formation of complex B (Fig.
8A, lanes 25-26).
By contrast, antibodies against octamer factors (Fig. 7A,
lanes 16-17, 19-20, 22-23), the
irrelevant FLAG motif (lanes 28-29), or zinc finger factors
BTEB2 (lanes 10-11), Gli (lanes 13-14), or CTCF
(lanes 7-8) factors that would potentially recognize the
CCTC motif had no effect on complex B formation. Anti-Fos (Fig.
7A, lanes 1-2) and anti-Jun (lanes
4-5) antibodies had weak effects on formation of the USF complex.
Notably, anti-USF1 did not perturb formation of the AP1 binding complex
forming on this region of the OPN promoter, indicating the
specificity of the immunoreagent (not shown). No nonspecific
interaction of antibody with radiolabeled probe was observed (Fig.
7A, lanes 3, 6, 9, 12, 15, 18, 21,
24, 27, and 30). Complex disruption without
supershift was noted with the anti-USF1 antibody (Fig. 7A).
Therefore, to provide additional evidence for endogenous USF
recognition of this OPN element, we probed the complex with
polyclonal antibodies directed against two completely different domains
of USF1. Both antibodies directed to a USF1 internal domain
(H86; Fig. 7B, lanes 1-3) and USF1
C-terminal domain (C20; Fig. 7B, lanes
4-6) inhibited complex B formation on OPN-(85/65). USF1 is
known to form a heterodimer in cells with the related family member
USF2 (54, 55). To provide further support that this complex B was
indeed USF, we tested the effects of anti-USF2 antibody on complex
formation. As shown in Fig. 7B, lanes 7-9, like
anti-USF1, anti-USF2 antibody disrupts formation of complex B. These
antibodies had no effect on the formation of DNA-protein complexes on
the AP1 cognate (not shown), indicating specificity of the
immunological USF1 reagents. Cold competition studies again confirm
that this very complex that is perturbed by anti-USF antibodies
recognized the CCTCATG motif in the OPN promoter, as
suggested in the data for Fig. 6. (Again, these gel shifts were carried
out in the presence of cold AP1 oligo to permit robust visualization of
complex B.) Unlike the wild type oligo sequence (Fig. 7B,
lanes 10-12), oligos with MUT #1 (lanes 13-15)
or MUT #2 (lanes 16-18) alterations that perturb the
CCTCATG E-box-like motif (Fig. 4) cannot compete for complex B
formation. By contrast, in MUT #3. (An alteration does not alter
this E-box-like cognate; Fig. 4) can still compete for formation of
complex B (Fig. 7B, lanes 19-22), consistent
with data presented above (see Fig. 6). Thus, the A7r5 protein-DNA complex B assembled by the proximal OPN promoter
region 85 to 65 is a heterodimer of USF1 and USF2 that recognizes
the CCTCATG motif at nucleotides 80 to 74.
View larger version (55K):
[in this window]
[in a new window]
|
Fig. 7.
Identification of USF1 and USF2 in the
protein-DNA complexes bound by the proximal OPN promoter
region 85 to 65. Gel shifts were carried out as described in
the legend to Fig. 6 but in the presence of cold AP1 cognate from the
MMP1 promoter (26) to preclude AP1 binding to the OPN
promoter fragment 85 to 65 and facilitate visualization of complex
B. A, A7r5 extracts analyzed by gel shift after
preincubation with 3 µg of the indicated antibodies. Lanes
1, 4, 7, 10, 13,
16, 19, 22, 25, and
28, basal binding. Lanes 2, 5,
8, 11, 14, 17,
20, 23, 26, and 29, binding in
the presence of the indicated antibody. Note that the anti-USF1
antibody (lane 26) markedly diminished formation of this
protein-DNA complex. Anti-Fos anti-body also slightly reduced complex
formation as well. No nonspecific interaction of antibody with
radiolabeled probe was observed (lanes 3,
6, 9, 12, 15,
18, 21, 24, 27, and
30). B, antibodies directed to the mid-region
(H86, lanes 1-3) or C terminus (C20,
lanes 4-6) of USF1 inhibit formation of complex B, as do
antibodies against USF2 (lanes 7-9). Competition assays
(lanes 10-22) confirm the binding specificity of this
USF1:USF2 complex is that of complex B. See Fig. 4 for competitor
sequences.
|
|
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 8.
Identification of c-Fos and phospho-c-Jun in
protein-DNA complexes bound by the proximal OPN promoter
region 85 to 65. Immunological probing of protein-DNA
complexes assembled with specific Fos and Jun family members.
Autoradiogram exposure times that reveal the significant supershifted
complexes (e.g. lanes 2, 17, and 21)
obliterate the resolution of complexes A, B, and C. Lanes 1,
4, 7, 13, 16, and
19, basal binding; lanes 2, 5,
8, 11, 14, 17, and 20,
binding in the presence of the indicated antibodies. Note that an
anti-cFos antibody disrupts and partially supershifts the complex
(lane 2). Notably, a small amount of basal binding activity
remains after treatment with anti-c-Fos antibody, reflecting the
residual USF binding activity unperturbed by c-Fos antibody (lane
2). Antibodies to other Fos family members did not affect binding
(lanes 4-12). Antibody to c-Jun reduced basal binding
(lane 14). Moreover, antibody to the activated phospho-c-Jun
supershifted the complex (lane 17). A weaker supershift is
observed with antibody to JunD (lane 20). None of the
antibodies interacts with the radiolabeled OPN probe in the
absence of extract (lanes 3, 6,
9, 12, 15, 18, and
21).
|
|
A7r5 Aortic VSMC DNA Binding Activities Containing c-Fos, c-Jun,
and JunD Recognize the OPN Promoter Region 85 to 65--
The
competition studies presented above identified that bands C and A
represent complexes assembled by the OPN promoter region 85 to 65
contain factors recognizing AP1 cognates. To provide additional
evidence for AP1 protein-DNA interactions and to identify the specific
VSMC AP1 complexes binding the OPN promoter, we
immunologically probed these complexes as described previously under
"Experimental Procedures." As shown in Fig. 8, antibody to c-Fos
markedly diminished and partially supershifted complexes formed on
radiolabeled OPN-(85/65) (Fig. 8, lanes 1-2).
Autoradiogram exposure times that reveal the significant supershifted
complexes (e.g. Fig. 8, lanes 2, 17,
and 21) obliterate the resolution of complexes A, B, and C. Notably, a small amount of basal binding activity remained after treatment with anti-c-Fos antibody, presumably reflecting the residual
USF binding activity unperturbed by c-Fos antibody (Fig. 8, lane
2). By contrast, antibodies to other Fos family members had no
effect on complex formation (Fig. 8, lanes 4-12). AP1
family members are typically heterodimers of specific Fos and Jun
proteins (51). As shown in Fig. 8, antibodies to c-Jun and
phospho-c-Jun partially reduced and supershifted the complex (Fig. 8,
lanes 13-17). A subset of the AP1 complexes assembled by
the OPN promoter contain JunD, revealed again by immunological
supershift (Fig. 8, lanes 19-20). None of the antibodies
interacted nonspecifically with the radiolabeled oligos
(Fig. 8, lanes 3, 6, 9,
12, 15, 18, and 21).
Antibodies to CTCF, BTEB2, Gli1, Oct1, Oct2, and Oct4 do not perturb or
supershift any of the complexes assembled by the OPN promoter 85 to
65 (not shown). Thus, AP1 binding factors recognize and regulate the
OPN promoter region 85 to 65. c-Fos, active
(phosphorylated) c-Jun, and JunD are present in the A7r5 aortic VSMC
complexes assembled by this region of the OPN promoter that
supports basal and glucose-dependent activity in A7r5 cultures.
The OPN Promoter USF Binding Activity in A7r5 Aortic VSMCs Is
Regulated by Glucose and 2-DG--
We next examined the effects of
glucose and 2-DG treatment on formation of the USF:OPN and
AP1:OPN protein-DNA interactions. Extracts were prepared
from A7r5 aortic VSMCs treated with either vehicle (control; 0 mM glucose), 5 mM glucose, or 5 mM
mannitol as described under "Experimental Procedures." As shown in
Fig. 9A, 5 mM
glucose treatment increased the formation of the AP1 and USF binding
activities that regulate OPN promoter. However, 5 mM mannitol treatment has no effect (USF binding again
visualized in presence of excess cold AP1 oligo; see above). The
influence of 2-DG in the presence of 5 or 30 mM glucose was
also examined. As shown in Fig. 9B, treatment with 2-DG did
not significantly alter total AP1 binding (Fig. 9B,
upper panel). However, treatment with 2-DG markedly
down-regulated the USF binding activity recognizing the OPN
promoter region 85 to 65 in the presence of 5 mM
glucose (Fig. 9B, lanes 3-6 versus lanes
1-2, lower panel; USF binding again was visualized in
the presence of excess cold AP1 oligo). Moreover, the addition of 30 mM glucose completely prevented the down-regulation of USF
binding to the OPN promoter by 2-DG (Fig. 9B, lower panel, lanes 9-10). Thus,
glucose and 2-DG regulate USF and AP1 binding activities, recognizing
the proximal OPN promoter region 85 to 65 in A7r5 VSMCs.
View larger version (38K):
[in this window]
[in a new window]
|
Fig. 9.
Glucose regulates AP1 and USF DNA binding
activities in A7r5 aortic VSMCs. A, extracts were prepared
from A7r5 cells treated with vehicle, 5 mM glucose, or 5 mM mannitol at the indicated concentrations for 1 day
(independent duplicates), and DNA binding activities recognizing the
OPN promoter region 85 to 65 were identified by gel shift assay.
Upper panel, total binding activity, reflecting primarily
AP1 (see Fig. 9); lower panel, USF binding activity assessed
as described in the legend to Fig. 8. Note that glucose up-regulates
both activities, whereas 5 mM mannitol does not.
B, regulation of binding by 2-DG. Upper panel,
total AP1 binding activity (see Fig. 9); lower panel, USF
binding activity. Note that 2-DG down-regulates USF binding activity
recognizing the OPN promoter (lanes 3-6) and that 30 mM glucose (Glc) reverses the suppression of USF
binding by 2-DG treatment (lanes 7-8). See "Results"
for details.
|
|
USF Protein Accumulation Is Regulated in Cultured Aortic VSMCs by
Glucose and 2-DG--
USF binding was markedly sensitive to glucose
and 2-DG treatment. We wished to identify whether these changes in USF
binding activity reflect changes in the accumulation of USF1 and/or
USF2 protein levels. Therefore, we performed Western blot analyses of
extracts prepared from A7r5 cells treated with either vehicle (0 mM glucose) or 5 mM glucose or mannitol. As
shown in Fig. 10A, treatment
of cells with glucose up-regulated USF1 but not USF2 protein
accumulation (independent duplicates). By contrast, mannitol exerted
little if any effect on USF protein accumulation. The effect of 2-DG in
the presence of 5 (basal) or 30 mM glucose was also
examined independently (Fig. 10B). As shown, 2-DG weakly
down-regulated USF1 protein levels in A7r5 VSMCs (Fig. 10B,
lanes 2 and 3 versus 1); by contrast,
30 mM glucose up-regulated USF1 protein accumulation and
completely prevented the 2-DG-mediated decrements in protein accumulation (Fig. 10B, lanes 4-5). USF2 levels
are not regulated by manipulation of glucose tone in A7r5 VSMCs (Fig.
10B), indicating specificity for regulation of USF1 protein
accumulation. Similar results were obtained in cultures of primary
aortic mesenchymal cells (not shown). Thus, USF1 protein accumulation
is specifically responsive to manipulation of cellular glucose tone in
A7r5 aortic VSMCs and primary mouse aortic VSMCs, consistent with
effects on total OPN USF binding activity. Regulation of the
OPN-(80/72) promoter USF binding activity in culture A7r5 aortic
VSMCs by 2-DG and glucose reflects in part the regulation of USF1
protein accumulation.
View larger version (42K):
[in this window]
[in a new window]
|
Fig. 10.
Regulation of USF1 protein levels in A7r5
cells by glucose. A, A7r5 aortic VSMCs were cultured in
glucose-free DMEM with 10% dialyzed FCS supplemented either with
vehicle or the indicated concentrations of glucose or mannitol
(independent duplicates) as described in the legend to Fig. 9. Cell
extracts were prepared and assays for USF protein accumulation by
Western blot analyses. Note that glucose up-regulates USF1 but not USF2
protein accumulation. Further note that 5 mM mannitol does
not regulate USF protein levels. B, extracts were prepared
from A7r5 VSMCs treated with vehicle (lane 1), 2 mM 2-DG (lane 2), 4 mM 2-DG
(lane 3), 30 mM glucose (lane 4), or
2 mM 2-DG with 30 mM glucose (lane
5). USF1 and USF2 protein levels were assessed by Western blot.
Note that glucose increases USF1 protein levels (lane 4 versus 1) and prevents 2-DG suppression (lane 5 versus 2). USF2 levels are not significantly
regulated.
|
|
The OPN Promoter Is Regulated in Transient Co-transfection Assays
by USF1 and AP1 (c-Fos:c-Jun) in A7r5 Aortic VSMCs--
To examine the
effects of AP1 (c-Fos:c-Jun) and USF (USF1:USF2) on OPN
promoter activity, we transiently co-transfected eukaryotic expression
vectors for c-Fos, c-Jun, USF1, and USF2 with the OPNLUC reporter in
A7r5 aortic VSMCs. As shown, both USF (Fig.
11A) and AP1 (c-Fos:c-Jun;
Fig. 11B) significantly augmented OPN promoter activity in A7r5 cells. USF regulation is complex, because mutation of
the composite cognate at 80 to 72 did not completely prevent OPNLUC
activation (not shown); this likely reflects the well described capacity of USF to activate transcription both via E-box elements and
also via the pyrimidine-rich initiator region element (Inr; CYCAYNYN)
(56-58) common to a number of promoter start sites (59, 60) including
the OPN gene (22). Synergy for OPN promoter activation
between AP1 and USF was not observed (Fig. 11A). USF dominated and limited AP1-dependent activation of the
OPN promoter (Fig. 11), and a 10-fold greater expression of
AP1 with USF resulted in transcriptional squelching, suggesting the
sequestration of an unidentified, rate-limiting co-regulator by these
factors (not shown; see "Discussion"). AP1 (c-Fos and c-Jun)
transactivates the proximal OPN promoter 83 to +78, and
activation was completely abrogated by mutation of the CCTCATGAC
element in the OPN promoter region 80 to 72 (MUT
#2; Fig. 11B). Thus USF and c-Fos:c-Jun both regulate
the proximal OPN promoter and function to support OPN gene
transcription in A7r5 rat aortic VSMCs.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 11.
USF and AP1 (c-Fos:c-Jun) up-regulate basal
OPN promoter activity in A7r5 aortic VSMCs. A7r5 cells
were co-transfected with 83 OPNLUC and eukaryotic expression vectors
for USF subunits (USF1, USF2) or AP1 subunits (c-Fos, c-Jun) as
indicated. Empty expression vector adjustment was used to maintain
constant DNA concentrations in all transfections. A,
transfection of USF (USF1/USF2) significantly up-regulates proximal
OPN promoter activity in A7r5 cells. No additive or
synergistic effects were observed with AP1 expression. B,
AP1 expression (c-Fos, c-Jun) significantly up-regulates basal
OPN promoter activity. Mutation of the CCTCATGAC motif to
CCTCATTAT (MUT #2) at 80 to 72 completely abrogates the
AP1 response. See "Results" for details.
|
|
Transactivation Directed by c-Fos and USF1 Is Augmented by Glucose
in A7r5 Aortic VSMCs--
In addition to DNA binding activity,
transcriptional activation function can be regulated by signaling
cascades that control functionally important protein-protein
interactions with the basal transcriptional machinery. We wished to
assess whether the transactivation functions AP1 and USF1, the factors
recognizing and regulating glucose-responsive OPN proximal
promoter, were regulated by glucose, as further evidence for their role
in transcriptional responses to glucose in A7r5 cells. Therefore, we
performed one-hybrid analyses, assessing the ability of c-Fos, c-Jun,
and USF1 to convey glucose responsiveness onto the unresponsive Gal4
DNA-binding domain (G4DBD) using five copies of the Gal4 response
element placed upstream of the LUC reporter gene (Gal4RE-LUC or
pFRLUC). As shown in Fig. 12A, the Gal4 DNA-binding
domain (which lacks a transcriptional activation domain) does
not activate transcription driven by the Gal4RE-LUC either in the
presence or absence of glucose treatment. By contrast, basal activity
was significantly augmented by expression of the Gal4DBD-c-Fos fusion
protein. Moreover, transcription was further increased by treatment
with 50 mM glucose (Fig. 12A). Of note,
Gal4DBD-cJun only weakly up-regulated basal activity and is minimally
responsive to glucose. Gal4DBD-USF1 also increased basal activity
weakly. However, the transactivation function of Gal4DBD-USF1 is
up-regulated 11-fold by 50 mM glucose treatment (Fig.
12A). Transactivation was also augmented at lower glucose concentrations (Fig. 12B). Moreover, as observed in the
OPN promoter context, c-Fos and USF1 transactivation were
enhanced by 5 mM glucose but not by 5 mM
mannitol (Fig. 12B). Thus, transactivation functions of
c-Fos and USF1, the DNA binding factors that recognized the
glucose-responsive OPN proximal promoter region, are
selectively activated by glucose treatment of A7r5 aortic VSMCs.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 12.
Transactivation functions of c-Fos and USF1
are augmented by glucose in A7r5 aortic VSMCs. A7r5 cells were
co-transfected with Gal4RE-LUC (pFRLUC) and pFACMV expression vectors
encoding the indicated Gal4 DBD (G4DBD) fusion proteins.
Transcription driven from Gal4RE-LUC was assayed from cells
treated either with vehicle (control) or 50 mM glucose.
A, note that glucose specifically augments c-Fos and
USF1-dependent transcription. B,
further note that 5 mM glucose, but not 5 mM
mannitol, induces c-Fos and USF1 transcriptional activation. See
"Results" for details,
|
|
| |
DISCUSSION |
OPN is a multifunctional extracellular protein that mediates
attachment and migration of osteoblasts, osteoclasts, macrophages, and
mesenchymal VSMCs (14). OPN enhances activation of cells of the
monocyte/macrophage lineage via up-regulation of interleukin-12 and
suppression of interleukin-10, thus recapitulating a Th1-type cytokine
activation profile (14). One cell type derived from the
monocyte/macrophage lineage is the osteoclast. Elegant studies by
Denhardt, Noda, and colleagues (61) have now identified a role for OPN
in osteoclast-dependent bone resorption activated by
estrogen deficiency parathyroid hormone treatment, RANKL (receptor activator of NF-
B ligand) stimulation (62), or mechanical unloading (63). Thus, in bone, OPN emerges as a crucial regulator of mineralized tissue turnover.
Unlike bone, the role of OPN in the arterial vasculature is poorly
characterized, and little is known of its transcriptional regulation in
this tissue. The OPN promoter is highly modular (14).
The OPN promoter can be organized into four regions:
(i) a nuclear receptor and Ras-responsive region (64) encoded by the
OPN promoter sequence 1000 to 550; peroxisome
proliferator-activated receptor 
(65), vitamin D receptor (44), and
estrogen receptor-related receptor- responses (66-68) all map to
elements in this region; (ii) a second region entraining expression to
the osteoblast differentiation program; BMP2-regulated Hoxc8:Smad1 (69)
and Runx2:Ets1 (46) transcriptional responses map to the OPN
promoter region 220 to 110; (iii) a basal promoter region that maps
to nucleotides 107 to +78 relative to the start site (see below); and
(iv) an intronic enhancer at +799 to +864 that supports Sox2:Oct4
regulated expression in pre-implantation embryonic development (70). It is likely that the Runx2 and BMP2 regulatory cascades play a crucial role in vascular OPN expression at later stages of calcific
vasculopathy (71), i.e. once the progressive, osteogenic
regulation of macrovascular calcification has been established as a
consequence of responses to initial metabolic or inflammatory insults
(20, 71-73). However, we identified that the OPN promoter
regions required for Smad responses (220 to 190) and Runx2:Ets
recognition (134 to 129; 118 to 113) in osteoblasts are not
required for basal or glucose-regulated OPN promoter
activity in A7r5 aortic VSMCs.
In our studies, we determined that the proximal OPN
promoter region 83 to 45 assembles DNA-protein interactions
sufficient for basal and acute glucose-regulated transcription in the
VSMC. We show that specific AP1 and USF complexes recognize and
regulate the OPN promoter via a CCTCATGAC element in this
region. Using rat VSMCs derived from either embryonic or adults
sources, Giachelli and co-workers (45) first identified that USF1 binds
the E-box cognate CAGGTG present at nucleotides 105 to 100 in the
murine OPN gene (see Fig. 4); USF2 association, regulation
by altered glucose signaling, and transcriptional responses to
transient co-expressed USF1 were not examined in their seminal study.
In data not shown, we also identified protein-DNA interactions at this
same E-box in A7r5 cells, confirming the results of Gia
TOP
ABSTRACT
INTRODUCTION
