| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Article |
From the Department of Internal Medicine and Cardiovascular Disease, Department of Biochemistry and Molecular Biology, Molecular Medicine Program, Mayo Clinic and Foundation, Rochester, Minn, and the Center for Thrombosis and Hemostasis (J.-Y.C.), University of North Carolina, Chapel Hill, NC.
Correspondence to Robert D. Simari, MD, Mayo Clinic and Foundation, GU 1801, 200 First St SW, Rochester, MN 55905. E-mail simari.robert{at}mayo.edu
Abstract
AbstractTissue
factor (TF) is a low-molecular-weight glycoprotein that
initiates the extrinsic clotting cascade and is considered a major
regulator of arterial thrombogenicity. TF pathway
inhibitor (TFPI) is a major
physiological inhibitor of TF-initiated
coagulation. The aim of this study was to define the complex interplay
between TF and TFPI and the regulation of vascular thrombogenicity in a
model of vascular remodeling. To determine the levels and pattern of
vascular expression of TF and TFPI associated with vascular remodeling,
a murine model of flow cessation was studied. TF activity of the
arteries increased after ligation
(P<0.05). Quantitative
analysis of homogenates of remodeled carotid
arteries revealed increased TF expression but unchanged TFPI expression
compared with normal carotid arteries, resulting in enhanced TF
activity. To determine the potential therapeutic role of TFPI in this
thrombogenic state, mice were treated with intravascular adenoviral
delivery of either murine TFPI (Ad-mTFPImyc) or a control adenovirus
(Ad-
E1). Overexpression of TFPI decreased vascular TF activity
compared with viral control
(P<0.01). Overexpression of
TFPI inhibited neointimal formation
(P=0.038), resulting in
enhanced luminal area (P=0.001)
4 weeks after flow cessation. In this murine model of vascular
remodeling, an imbalance between TF and TFPI expression is generated,
resulting in increased TF activity. Overexpression of TFPI in this
model inhibits vascular TF activity and results in attenuation of
vascular remodeling associated with flow interruption.
Key Words: thrombosis arteriosclerosis gene therapy
Arterial thrombosis is the proximate cause of myocardial infarction and stroke. Biochemical and clinical evidence suggests the importance of the thrombogenic nature of atherosclerotic vessels in this process. Tissue factor (TF), a low-molecular-weight glycoprotein that initiates the extrinsic clotting cascade, is considered a major regulator of coagulation, hemostasis, and thrombogenicity of atherosclerotic arteries.1 2 3 4 5 6 7 8 9 TF pathway inhibitor (TFPI), which provides physiological inhibition of TF-initiated coagulation by binding to factor Xa and the TFfactor VIIa complex in a 2-step process, is found in vascular endothelium and smooth muscle cells as well as in platelets, blood monocytes, and macrophages.10 11 12 13 14 In atherosclerotic carotid arteries, TF expression is abundant, whereas TFPI expression is limited in up to 30% of plaques, resulting in predominant TF activity.15 In plaque, where TFPI expression is the greatest, TF activity is attenuated. Thus, this imbalance between TF and TFPI expression in plaque may result in the prothrombotic phenotype associated with atherosclerosis. To investigate the development and potential regulation of this imbalance, a well-defined murine model of vascular remodeling, including neointimal formation, was studied.16
Materials and Methods
Animal Model
The murine model of vascular remodeling associated
with carotid flow cessation as described by Kumar and
Lindner16 was used. All
procedures complied with the standards for care and use of animal
subjects as stated in the Guide for the
Care and Use of Laboratory Animals (Institute of Laboratory
Animal Resources, National Academy of Sciences, Bethesda, Md). Briefly,
via a ventral longitudinal incision, the left common carotid of adult
C57BL/6 mice was identified and ligated with 4-0 silk just proximal to
the bifurcation. The midline incision was closed with 6-0
vicryl, and the skin was closed with a
Nexaband topical skin closure kit supplied by
Veterinary Products Laboratories. Warmed lactated Ringers
solution (1 mL SC) and Enrofloxacin (Bayer) (0.01 mL IM) were
injected, and the mice were allowed to recover on a warm hydrothermal
pad. At 1, 2, 3, and 4 weeks, the ligated carotid arteries were
harvested as fresh frozen specimens for protein analysis, or
they were harvested after perfusion fixation at
physiological pressure with 10% formalin before
euthanasia.
Tissue Processing and Analysis
The carotid arteries were obtained from mice,
thoroughly rinsed in saline, and freshly frozen in liquid nitrogen.
Thawed tissue was subsequently homogenized in 50
mmol/L Tris HCl, pH 8.0, by using a tissue homogenizer.
The homogenate then was analyzed for protein
content by using a Bradford assay
(Bio-Rad).
Computer-aided morphometric analysis (Image Pro Plus) was performed on hematoxylin and eosinstained sections by two investigators blinded to the treatment. Briefly, after perfusion fixation and rinsing, the entire carotid artery was embedded and cut at 250-µm intervals from the ligature. Because of the nature of the remodeling process, contiguous sections starting at a fixed distance from the ligature after fixation and embedding (500 µm) were analyzed morphometrically. These sections were uniformly free of thrombus. The areas subtended by the external and internal elastic laminae and lumen were measured, and means were compared between the treatment groups.
Adenovirus Generation
A first-generation recombinant adenoviral vector
encoding for murine TFPI (Ad-mTFPImyc) was generated by using standard
techniques.17 Briefly, the
cDNA for murine TFPI (mTFPI) was kindly obtained from Dr George Broze
at Washington University, St. Louis, Mo. A
myc tag was placed on the 3'
end of the coding region to provide a unique epitope for detecting
transgene expression. The mTFPImyc cassette containing a
cytomegalovirus promoter/enhancer was cloned into a shuttle plasmid
containing LoxP sites and recombined with a cosmid containing
the Ad5 genome in the presence of Cre recombinase. This solution was
then transfected into 293 cells, resulting in recombinant adenoviral
generation. Final preparations were purified by using standard
doublecesium chloride banding and titered as
described.18 A control
adenoviral vector without cDNA insert (Ad-
E1) was used as a control
for adenoviral infection, and an adenovirus expressing
Escherichia coli
ß-galactosidase from the cytomegalovirus promoter/enhancer (Ad-LacZ)
was used as a reporter
construct.18
Immunohistochemical Staining for
ß-Galactosidase
Five-micron-thick transverse frozen sections of mouse
carotid arteries harvested 2 days after delivery of Ad-LacZ were fixed
with acetone for 10 minutes at -20°C, washed with PBS, and blocked
for 30 minutes with 10% normal donkey serum. The sections were then
incubated 1 hour with rabbit antiß-galactosidase antibody
(Molecular Probes) at a concentration of 10
µg/mL, washed with PBS containing Triton and normal donkey serum, and
incubated with a biotinylated donkey anti-rabbit antibody
(Amersham Pharmacia Biotech) for 30 minutes.
After a PBS wash containing Triton and normal donkey serum, the
sections were incubated for 45 minutes with streptavidinalkaline
phosphatase (Vector Laboratories). The reaction product was
visualized by using Vector Blue with levamisole (Vector Laboratories).
Nontransduced arteries failed to show immunoreactivity when studies
were performed in parallel. The rabbit anti-mTFPI antiserum was
prepared commercially (Cocalico Biologicals) after inoculation of
rabbits with 1 mg recombinant mTFPI made in eukaryotic
cells. Preimmunization serum was used as a control. A rooster
anti-serum to murine TF was used (gift from Dr Robert Kelm, Mayo Clinic
and Foundation, Rochester, Minn).
Adenoviral Delivery
To evaluate the effect of overexpression of TFPI on
vascular TF activity, an adenoviral vector generated as described above
was used to overexpress TFPI in the mice carotid arteries at the time
of carotid ligation. Carotid arteries were exposed in mice via a
ventral incision. The first suture was placed at the bifurcation of the
common carotid artery, and a second temporary suture was placed
1 cm
proximal to the first. Ten microliters of the viral solution containing
either 1.33x1010 plaque-forming units
(pfu)/mL (a total dose of 1.33x108 pfu) of
Ad-LacZ, Ad-mTFPImyc, or Ad-
E1 (without cDNA insert) was instilled
between these two sutures with use of a 33-gauge needle distending the
vessel. The needle was withdrawn, and the first suture placed at the
bifurcation was tied. The puncture site in the vessel was at such a
location that the first suture closed it. Twenty minutes later, the
second suture was released. The incision was closed with staples, and
the mice were allowed to recover.
Immunoblot Analysis
TFPI protein was analyzed in ligated
carotid arteries obtained at 1, 2, and 4 weeks. Equal amounts of
protein were denatured by boiling for 5 minutes and resolved by
electrophoresis on a 9% SDS-polyacrylamide gel. Transfer of
protein to a nitrocellulose membrane was carried out for 3 hours at
4°C. Immunoblotting was performed by using the
above-described rabbit anti-mouse TFPI antiserum at 1:500 dilution in
nonfat milk/Tris-buffered saline. After washes, the membrane was probed
subsequently with a rabbit secondary antibody conjugated to horseradish
peroxidase (Amersham Life Sciences) at 1:500
dilution and developed with chemoluminescence (Supersignal,
Pierce). The membrane then was exposed to x-ray
film (Kodak) and subsequently developed. Immunoblotting
for TF expression was performed by using rooster anti-mouse TF
antiserum at 1:500 dilution in nonfat milk/Tris-buffered saline and by
subsequently probing with a rabbit secondary antibody conjugated to
horseradish peroxidase. The tissue homogenates from the
mice treated with Ad-mTFPImyc or Ad-
E1 (without cDNA insert) were
also immunoblotted with monoclonal
anti-myc antibodies to
differentiate between native and adenovirus-induced
TFPI.
Determination of TF Activity
TF activity was determined as described
previously.19 Diluting the
samples with Tris HCl equalized the protein concentration in the
samples. Thirty microliters of sample was added to a fibrin cup in a
BBL Fibrosystem Fibrometer (Becton Dickson). This was followed
by the addition of 30 µL of rabbit brain cephalin diluted 1:10 in
0.85% NaCl and 30 µL of 0.02 mol/L CaCl2.
Thirty microliters of reconstituted mouse plasma was added immediately
thereafter. The fibrometer was started, and the clotting time was
recorded.
Statistical Analysis
Groups were compared by using unpaired Student
t tests. A value of
P<0.05 was considered
significant.
Results
Vascular Remodeling Results in Enhanced
Vascular Thrombogenicity
As previously described by Kumar and
Lindner,16 ligation of the
left carotid artery in C57BL/6 mice resulted in vascular remodeling
associated with neointimal formation. To determine whether
vascular remodeling results in increased vascular thrombogenicity,
vascular TF activity was determined by measuring murine plasma clotting
times in the presence of tissue homogenates from ligated
arteries at baseline and at 1, 2, and 4 weeks after surgery. There was
a significant decrease in the clotting time in the ligated arteries at
each time point after ligation
(P<0.05) compared with
baseline
(Figure 1
). These results suggest that vascular remodeling in
this murine model is associated with enhanced vascular
thrombogenicity.
|
Vascular Remodeling Leads to Imbalance Between
TF and TFPI
To define the mechanism of altered
thrombogenicity of vascular homogenates, the levels and
pattern of vascular expression of TF and TFPI associated with vascular
remodeling were studied. Immunoblotting for TF and TFPI
expression was performed in arterial
homogenates from ligated arteries at 1, 2, and 4 weeks
after surgery. Quantitative analysis of the
immunoblots revealed that TF expression was increased in
the ligated carotid arteries of every animal (n=3) at each time point
compared with normal carotid arteries
(Figure 2
). Expression of TF was maximal at 2 weeks after
flow cessation. In contrast, TFPI expression was unchanged in ligated
arteries compared with normal arteries. Thus, a pattern of gene
expression favoring a thrombogenic phenotype associated with
vascular remodeling was seen in this model as in previous studies of
human
tissue.15
|
Overexpression of Murine TFPI
To determine whether the enhanced TF activity can be
attenuated by heterologous vascular expression of TFPI, a
replication-deficient adenoviral vector expressing mTFPI was generated
(Ad-mTFPImyc). To overexpress mTFPI in ligated arteries, a direct in
vivo method of gene transfer was established. To confirm gene transfer,
immunostaining of arteries 2 days (at peak expression)
after delivery of Ad-LacZ demonstrated transgene expression in the
transduced arteries
(Figures 3A
and 3B
). With this technique of distending luminal
delivery, transgene expression was detected in cells within the
adventitial portions of the arteries.
|
To further define the ability to overexpress mTFPI in
ligated arteries, either Ad-mTFPImyc or Ad-
E1 was delivered at the
time of ligation, and the animals were euthanized at 1 week.
Immunoblotting revealed specific overexpression of the
mTFPImyc construct by enhanced immunoreactivity to anti-mTFPI antiserum
(Figure 3C
). To determine whether overexpression of mTFPI
would attenuate the enhanced TF activity associated with vascular
remodeling, TF activity was measured in the ligated arteries (n=5 per
group) treated with Ad-mTFPImyc or Ad-
E1 at 1 week. There was a 26%
increase (P<0.001) in the
homogenate plasma clotting time in the Ad-mTFPImyc group
compared with the viral control group. These data suggest that the
imbalance between TF and TFPI results in increased TF activity with
vascular remodeling and that this increase can be modulated by
enhancing vascular TFPI expression.
TFPI Overexpression Attenuates Vascular
Remodeling
To determine whether overexpression of TFPI resulting
in inhibition of vascular TF activity might affect vascular
remodeling in this murine model, the effects of Ad-mTFPImyc infection
on vascular remodeling were studied and compared with control
adenoviral infection (Ad-
E1). Four weeks after ligature, the
Ad-mTFPImyc group demonstrated less neointimal formation,
as measured by the neointimal/medial ratio (0.35±0.18
versus 1.7±0.50 in Ad-
E1,
P=0.038)
(Figure 4
). There was no difference between the groups in
external elastic laminae or medial areas. However, the luminal area was
greater in the Ad-mTFPImyc group compared with the viral control group
(2101±1254 µm2 for Ad-
E1 versus
15 482±2049 µm2 for Ad-mTFPImyc,
P=0.0014). Taken together,
these data suggest that the TF pathway plays a role in the development
of vascular remodeling in this murine
model.
|
Discussion
Human studies suggest that vascular thrombogenicity is regulated by vascular expression of prothrombotic and antithrombotic factors. TF plays a central role in this process6 7 8 in spite of vascular expression of its inhibitor, TFPI.15 20 There are three lines of evidence that support the potential for TFPI to modulate vascular TF activity. First, studies in explanted human tissue suggest that expression of TFPI is related to attenuated TF activity.15 Second, studies have shown that inherent or recombinant TFPI alters local vascular TF activity9 21 22 and neointimal formation.23 24 Finally, gene transfer studies of TFPI have resulted in local antithrombotic effects.25 26 However, insights into the establishment and local regulation of this imbalance between TF and TFPI have not been defined. To do so, an animal model that mimics important aspects of human vascular disease without acute endothelial denudation was used. Unlike endothelial injury models of remodeling, this model results in the development of neointima in the absence of direct endothelial injury as a result of changes in shear stress and stasis. Thus, this model attempts to model the pathophysiological situation of lesion development in the absence of noticeable endothelial denudation.27
A key finding of the present study is that the arterial wall developed a prothrombotic phenotype in this murine model concurrent with neointimal formation and vascular remodeling. This thrombogenic phenotype was due to increased vascular TF expression without a concomitant increase in TFPI expression. This imbalance resulted in increased vascular TF activity, which mirrors the imbalance demonstrated in human atherosclerotic plaque.15 Furthermore, vascular overexpression of mTFPI was performed by using a novel direct-delivery system. Overexpression of mTFPI attenuated the enhanced TF activity and attenuated vascular remodeling. Taken together, these data support the concept that this model of vascular remodeling is associated with a thrombogenic phenotype and that inhibition of TF activity by TFPI may inhibit vascular remodeling.
These data clearly identify a relative vascular deficiency of TFPI and an increase in TF in this model. The regulation of TF expression is well defined, and a number of stimuli for TF production in vascular cells have been identified, including thrombin, growth factors, and cytokines.28 29 In this model, TF is expressed within the media and the growing neointima from vascular and blood-borne sources.30 The lack of parallel increases in TFPI may be due to differences between the vascular regulation of TF and TFPI and the recruitment of nonvascular sources of TF.31 These differences favored robust TF expression in the setting of uninducible TFPI expression. Thus, the imbalance that was identified in the present study may be an ultimate result of these regulatory differences. Attenuation of TF activity by a TFPI transgene driven by viral regulatory sequences supports this contention.
The present study provides a mechanism to explain the results of two prior studies of TFPI gene transfer that demonstrated inhibition of vascular thrombosis in animal models.25 26 Distinct from these studies, the present study determines that TFPI overexpression is capable of modulating vascular TF activity and uses a species-matched form of the protein. The delivery technique used in the present study resulted in adventitial transgene expression, and the intimal effects seen likely represent the potent secreted nature of the TFPI transgene affecting intravascular and/or intraluminal TF activity. Thus, the ability of TFPI expressed in the vasculature to inhibit TF activity likely explains the previously defined effects of TFPI overexpression in distinct animal models of intravascular thrombosis. To further delineate whether the effects seen in the present study are due to intravascular or endoluminal expression, studies with genetically determined alterations in vascular, systemic, or blood-derived TFPI expression will be performed.
Perhaps the most interesting finding of the present study is the effect of TF inhibition on the structural response to flow cessation without endothelial denudation. Previous studies have identified the importance of growth factors32 and P-selectin16 in the development of remodeling in this model. There are several potential mechanisms for TF and TFPI as regulators of vascular structure. Because intravascular or intramural fibrin deposition may act as a scaffold for plaque growth, the TF pathway may play an indirect role. In addition, TF may act via thrombin and factor Xa as a mitogen for vascular smooth muscle cells.33 TFPI has been shown to inhibit vascular smooth muscle cells in vitro.34 A role for TF in cellular migration has been proposed.35 Thus, inhibition of TF may have multifactorial effects on vascular remodeling.
Taken together, the findings of the present study clearly establish that vascular remodeling results in a prothrombotic vascular phenotype. This prothrombotic phenotype is, at least in part, due to an imbalance between TF and TFPI, which results in increased vascular TF activity. Attenuation of this increased thrombogenicity and neointimal formation by overexpression of TFPI provides evidence for therapeutic possibilities of mitigating both the enhanced thrombogenicity and vascular remodeling associated with vascular disease.
Acknowledgments
This study was supported by grants from the National Institutes of Health (HL-03473-5 and HL-65191-1), the Bruce and Ruth Rappaport Program in Vascular Biology, the Miami Heart Research Institute, and the Mayo Foundation. Appreciation is extended to Traci Paulson for technical and secretarial support.
Footnotes
Original received February 22, 2001; revision received April 16, 2001; accepted May 2, 2001.
References
This article has been cited by other articles:
![]() |
N. Mackman and M. B. Taubman Does Tissue Factor Expression by Vascular Smooth Muscle Cells Provide a Link Between C-Reactive Protein and Cardiovascular Disease? Arterioscler. Thromb. Vasc. Biol., April 1, 2008; 28(4): 601 - 603. [Full Text] [PDF] |
||||
![]() |
N. Hostettler, A. Naggi, G. Torri, R. Ishai-Michaeli, B. Casu, I. Vlodavsky, and L. Borsig P-selectin- and heparanase-dependent antimetastatic activity of non-anticoagulant heparins FASEB J, November 1, 2007; 21(13): 3562 - 3572. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. S. BelAiba, T. Djordjevic, S. Bonello, F. Artunc, F. Lang, J. Hess, and A. Gorlach The Serum- and Glucocorticoid-Inducible Kinase Sgk-1 Is Involved in Pulmonary Vascular Remodeling: Role in Redox-Sensitive Regulation of Tissue Factor by Thrombin Circ. Res., March 31, 2006; 98(6): 828 - 836. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Pandolfi C-reactive protein: A potential new molecular link between inflammation, thrombosis and vascular cell proliferation? Cardiovasc Res, October 1, 2005; 68(1): 3 - 4. [Full Text] [PDF] |
||||
![]() |
L. V. M. Rao and U. R. Pendurthi Tissue Factor-Factor VIIa Signaling Arterioscler. Thromb. Vasc. Biol., January 1, 2005; 25(1): 47 - 56. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. M. Day, J. L. Reeve, B. Pedersen, D. M Farris, D. D. Myers, M. Im, T. W. Wakefield, N. Mackman, and W. P. Fay Macrovascular thrombosis is driven by tissue factor derived primarily from the blood vessel wall Blood, January 1, 2005; 105(1): 192 - 198. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. W. Kopp, T. Holzenbein, S. Steiner, R. Marculescu, H. Bergmeister, D. Seidinger, I. Mosberger, C. Kaun, M. Cejna, R. Horvat, et al. Inhibition of restenosis by tissue factor pathway inhibitor: in vivo and in vitro evidence for suppressed monocyte chemoattraction and reduced gelatinolytic activity Blood, March 1, 2004; 103(5): 1653 - 1661. [Abstract] [Full Text] [PDF] |
||||
![]() |
R.P. Brandes, S. Beer, T. Ha, and R. Busse Withdrawal of Cerivastatin Induces Monocyte Chemoattractant Protein 1 and Tissue Factor Expression in Cultured Vascular Smooth Muscle Cells Arterioscler. Thromb. Vasc. Biol., October 1, 2003; 23(10): 1794 - 1800. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. L. Myers, K. J. Harmon, V. Lindner, and L. Liaw Alterations of Arterial Physiology in Osteopontin-Null Mice Arterioscler. Thromb. Vasc. Biol., June 1, 2003; 23(6): 1021 - 1028. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. M. Day, D. Duquaine, L. V. Mundada, R. G. Menon, B. V. Khan, S. Rajagopalan, and W. P. Fay Chronic Iron Administration Increases Vascular Oxidative Stress and Accelerates Arterial Thrombosis Circulation, May 27, 2003; 107(20): 2601 - 2606. [Abstract] [Full Text] [PDF] |
||||
![]() |
X. Yin, C. Yutani, Y. Ikeda, K. Enjyoji, H. Ishibashi-Ueda, S. Yasuda, Y. Tsukamoto, H. Nonogi, Y. Kaneda, and H. Kato Tissue factor pathway inhibitor gene delivery using HVJ-AVE liposomes markedly reduces restenosis in atherosclerotic arteries Cardiovasc Res, December 1, 2002; 56(3): 454 - 463. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Kato Regulation of Functions of Vascular Wall Cells by Tissue Factor Pathway Inhibitor: Basic and Clinical Aspects Arterioscler. Thromb. Vasc. Biol., April 1, 2002; 22(4): 539 - 548. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. J. Shebuski and K. S. Kilgore Role of Inflammatory Mediators in Thrombogenesis J. Pharmacol. Exp. Ther., March 1, 2002; 300(3): 729 - 735. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Corti, V. Fuster, and J.J. Badimon Strategy for ensuring a better future for the vessel wall Eur. Heart J. Suppl., February 1, 2002; 4(suppl_A): A31 - A41. [Abstract] [PDF] |
||||
![]() |
H. Kato Regulation of Functions of Vascular Wall Cells by Tissue Factor Pathway Inhibitor: Basic and Clinical Aspects Arterioscler. Thromb. Vasc. Biol., April 1, 2002; 22(4): 539 - 548. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Circulation Research Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2001 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |