(Circulation. 2000;101:2000.)
© 2000 American Heart Association, Inc.
Basic Science Reports |
From the Laboratoire dHématologie (D.C., L.R., B.J.), Service de Cardiologie et Hémodynamique (T.M., C.B.), Service de Pharmacologie (R.B.), Service Commun dImagerie Interventionnelle (I.S.), and Laboratoire dAnatomie et de Cytologie Pathologiques C (P.R.), Centre Hospitalier Régional Universitaire and Faculté de Médecine, Lille, France; and Section of Cardiovascular Medicine, Yale University, New Haven, Conn (M.D.E.).
Correspondence to Brigitte Jude, MD, Laboratoire dHématologie, Hôpital Cardiologique, Boulevard du Professeur J. Leclercq, 59037 Lille Cedex, France. E-mail b-jude{at}chru-lille.fr
| Abstract |
|---|
|
|
|---|
Methods and ResultsWe studied TF expression in normally fed (n=16) and cholesterol-fed (2% for 6 weeks, n=16) rabbits. Animals were then randomized to receive intravenous bFGF (2.5 µg twice weekly for 3 weeks) or saline injections. TF expression was evaluated in mononuclear cells from arterial blood and in aortic sections by an immunohistochemical assay using a monoclonal anti-rabbit TF antibody (activator protein 1). Monocyte TF expression was increased by bFGF administration in both normal and hypercholesterolemic rabbits (129±45 versus 19±3 mU TF/1000 monocytes, P<0.05, and 31±12 versus 7±1 mU TF/1000 monocytes, P<0.005, respectively) and was further increased by stimulation of monocytes by endotoxin in vitro. TF expression was lower in hypercholesterolemic rabbits than in normal rabbits. In the media of the vascular wall, bFGF induced strong TF expression in normal rabbits and only weak TF expression in hypercholesterolemic ones.
ConclusionsThis study demonstrates that systemic administration of bFGF induces an impressive increase of TF expression in circulating monocytes and in the vascular wall in normal and to a lower extent in hypercholesterolemic rabbits. The significance of this observation in terms of inducing thrombosis in vivo needs clarification.
Key Words: growth substances leukocytes vessels
| Introduction |
|---|
|
|
|---|
Of particular interest, basic fibroblast growth factor (bFGF) can
stimulate proliferation, morphological changes, and migration of
vascular smooth muscle cells (SMCs) and vascular
endothelial cells in vitro; in vivo, bFGF has been
shown to promote reendothelialization with functional
endothelium after balloon injury or induced
atherosclerosis.2 3 4 5 Although these
findings are advantageous, numerous genes can be activated by
growth factors, including bFGF, that may produce undesired effects. For
instance, tissue factor (TF) probably plays an important role in both
thrombogenesis6 and angiogenesis.7 8 9 The
regulation of the TF gene is controlled by several transcription
factors activated by inflammatory cytokines
(interleukin-1ß and tumor necrosis factor-
), oxidized LDLs, and
endotoxin.10 11 In vitro growth factors, such as
platelet-derived growth factor, FGF, transforming growth
factor-ß, or epidermal growth factor, can induce TF expression in
fibroblasts and SMCs.12 13 14 15
To investigate whether bFGF administration upregulates TF expression in the vasculature, we evaluated TF expression in monocytes and in the vascular wall in normal and hypercholesterolemic rabbits treated with bFGF.
| Methods |
|---|
|
|
|---|
In the normocholesterolemic and the hypercholesterolemic rabbits assigned to bFGF treatment, human recombinant bFGF boluses (2.5 µg bFGF in 1 mL 0.5% albumin per IV injection, Sigma) were administered twice weekly during a 3-week period. The rabbits assigned to placebo treatment received vehicle injections of albumin twice weekly.
Blood was obtained 3 days after the last injection of bFGF or placebo. All animals were then euthanized with sodium pentobarbital for histological studies. All experiments were conducted in compliance with the position of the American Physiological Society on research animal use.
Blood Samples
Blood was sampled under sterile conditions from the ear artery:
5 mL in lithium heparin (143 USP units) for mononuclear cell isolation,
2 mL in sodium citrate (1:10, 3.8%) for coagulation and lipid
analysis, and 1 mL in EDTA for blood cell counts.
Samples collected on sodium citrate were centrifuged (1500g, 15 minutes, +4°C) and separated. The plasma was frozen (-80°C) for further analysis.
Samples collected on EDTA were used for hematological analyses with a Coulter MAXM. The white blood cell counts were verified manually. Peripheral blood smears for differential white cell counts were stained with May-Grünwald-Giemsa stain and examined by 3 investigators who were blinded to the treatment allocation.
Mononuclear Cell Culture
The mononuclear cells were isolated by gradient
centrifugation (MSL, d=1.077±0.001,
Eurobio), washed twice, and resuspended in RPMI 1640 (Gibco)
(3x106 cells/mL). Monocytes made up 12±1%
(mean±SEM) of the cells. Cell viability was >98% (trypan blue
test).
All reagents and culture supplies used were free of endotoxin (chromogenic limulus amoebocyte lysate assay, sensitivity 0.025 endotoxin unit [EU]/mL). An aliquot of the freshly isolated mononuclear cells, referred to as noncultured cells, was frozen at -80°C.
Aliquots of cell preparations (3x106 cells/mL) suspended in RPMI 1640 without FCS were cultured for 16 hours at 37°C in a humidified 5% CO2 atmosphere, without or with stimulation by endotoxin (5000 EU/mL, Escherichia coli 055:B5, Sigma); these are referred to as unstimulated and stimulated cells, respectively. At the end of the incubation period, cells were resuspended and frozen at -80°C.
TF Activity Assay
The frozen cells were lysed by addition of 0.05 mol/L Tris/HCl,
0.1 mol/L NaCl, 0.1% Triton X100, 0.1% BSA (60 µL/mL) for 30
minutes at 37°C associated with serial vortex mixing. TF activity was
determined by a modified amidolytic assay as previously
described.16 Briefly, lysed cell suspensions (50 µL)
were incubated at 37°C mixed with CaCl2 0.25
mol/L (50 µL) and prothrombin concentrate complex (Laboratoire de
Fractionnement et des Biotechnologies) as a source of factor VII (50
µL, 3 IU/mL). After addition of 50 µL of the
chromogenic substrate S2765 (Biogenic), the change in
optical density (410 nm) was quantified and converted to units of TF
activity from serial dilutions of a rabbit brain thromboplastin (CI+,
Stago). One milliliter of thromboplastin was assigned a value of 1000
U/mL of TF. Results were expressed as mU/1000 monocytes.
The procoagulant activity was characterized as TF by a neutralization procedure using mouse monoclonal antibody anti-rabbit TF (AP-1, gift from M.D. Ezekowitz, Yale University, New Haven, Conn); diluted (1/18) antibody (25 µL) was incubated with diluted TF standard or lysed cell suspensions for 30 minutes at 37°C. Then the mixture was tested for procoagulant activity. The percentage neutralization was calculated with the rabbit brain thromboplastin without and with antibody used as 0% and 100%, respectively.
Additional Assays on Plasma Samples
Prothrombin time and fibrinogen levels were measured by clotting
assays (Biomerieux).
Factor II, V, and VII+X levels were determined by an automated clotting assay (STA, Stago) using calcified rabbit thromboplastin and human factordeficient plasma (Stago). Coagulation times were then expressed as percentages by comparison to standard curves constructed with serial dilutions of standard rabbit plasma in Owren Koller buffer (1/10 to 1/40 for factor II and factor VII+X and 1/100 to 1/400 for factor V).
Histological Studies
The abdominal aortas of all animals were dissected and cleaned
by brief immersion in PBS, fixed in 4%
paraformaldehyde for 10 minutes, and then maintained in
30% sucrose/distilled water for 4 hours. Tissues were embedded in OCT
(methyl methacrylate) compound, quick-frozen in isopentane, and stored
at -80°C. Sections (6 µm) of the frozen tissues mounted on
glass slides (Superfrost Plus) were immunohistochemically labeled with
a murine monoclonal antibody against rabbit TF (AP-1). Briefly, frozen
sections were air-dried for 1 hour, then blocked with 10% horse serum
for 10 minutes at room temperature. TF antibody was diluted at 1/200 in
PBS and incubated for 1 hour at 37°C. In parallel, a negative control
without primary antibody was performed. Samples were washed 3 times
with PBS and incubated with a biotinylated anti-mouse IgG (Vectastain
ABC kit, Vector Laboratories) diluted in PBS for 1 hour at room
temperature. Samples were washed 3 times with PBS, incubated with ABC
reagent for 1 hour, and color-developed for peroxidase with
diaminobenzidine (12 µL
H2O2, 15 mL PBS) for 3
minutes. Samples were counterstained with hematoxylin, dehydrated in
graded ethanols, and mounted with Eukitt (Pertex, Histolab). Two
independent investigators who were blinded to the treatment allocation
assigned semiquantitative scores to the slides (0 corresponding to no
TF staining, and 3 to strong TF staining, corresponding to adventitia
staining).
Statistical Analysis
Results are expressed as mean±SEM. Data were analyzed
with a nonparametric test (Kruskal-Wallis) to determine
significant differences (P<0.05) in means between groups,
followed by a Mann-Whitney U test to test the significance
of differences between groups.
| Results |
|---|
|
|
|---|
|
Monocyte TF Activity
In all groups, procoagulant activity was detectable in mononuclear
cells after 16 hours of culture. A neutralization assay with TF
antibody confirmed that procoagulant activity was due to TF expression
in all cases. The levels of TF activity in stimulated and unstimulated
cells in the 4 groups are shown in Figure 1
.
|
Effect of bFGF Administration in Normal Rabbits
TF activity was higher in the bFGF group than in the control
group. This difference was significant in unstimulated cells (129±45
versus 19±3 mU/1000 monocytes, P<0.05) but was not
significant in stimulated cells (124±38 versus 50±11 mU/1000
monocytes, P=NS).
Effect of Hypercholesterolemia
TF activity in unstimulated cells was significantly lower in the
hypercholesterolemic control group than in the control
group (7±1 versus 19±3 mU/1000 monocytes, P<0.05). TF
activity in stimulated cells was the same in the 2 groups (53±12
versus 50±11 mU/1000 monocytes, P=NS).
Effect of bFGF Administration in Hypercholesterolemic
Rabbits
TF activity was higher in the hypercholesterolemic
bFGF group than in the hypercholesterolemic control
group. This difference was significant in unstimulated cells (31±12
versus 7±1 mU TF/1000 monocytes, P<0.005) and in
stimulated cells (185±26 versus 53±12 mU TF/1000 monocytes,
P<0.005). However, this increase was not as high as it was
in the normocholesterolemic rabbits treated with bFGF
(31±12 versus 129±45 mU TF/1000 monocytes, P<0.05).
Histological Studies
Abdominal vessels were studied by immunohistochemistry to map the
TF distribution in the vessel wall. Representative
photomicrographs are shown in Figure 2
.
|
TF positivity was found predominantly in the adventitia in all groups.
There was no TF expression in the intima or the media of control
rabbits (Figure 2a
). In rabbits treated with bFGF, TF was found
in the intima and in the media. Endothelial staining
was particularly evident (Figure 2b
). In
hypercholesterolemic control rabbits, minimal amounts
of TF were detected in the media but none in the intima (Figure 2c
). After bFGF treatment in
hypercholesterolemic rabbits, a moderate increase of TF
expression was seen in the media but not in the intima (Figure 2d
). It is noteworthy that induction of TF expression by bFGF
was not as high in cholesterol-fed rabbits as in normally
fed rabbits (Figure 2d
and 2b
).
Changes in Other Parameters
In normal rabbits (Table 2
), bFGF
induced a decrease of the factor VII+X level and an increase of the
fibrinogen level and platelet count. In the
hypercholesterolemic groups, however, factor II, V, and
VII+X levels and platelet counts increased and fibrinogen levels
decreased with no further effect of bFGF.
|
| Discussion |
|---|
|
|
|---|
Effect of bFGF Administration in Normal Rabbits
The induction of TF gene by growth factors has already been
demonstrated in fibroblasts and SMCs. Our results indicate that bFGF
induces TF expression in vivo in monocytes and in vascular cells. Four
types of FGF receptors have been identified; these are present in
either monocytes/macrophages, endothelial
cells, or SMCs.17 18 19 The binding of bFGF to its
specific receptor triggers a cascade of events leading to signal
transduction by extracellular signal-regulated kinase (ERK),
mitogen-activated protein kinase, Ras, and protein kinase C
(PKC) pathways.20 21 Little is known about the induction
mechanisms of TF expression by bFGF. However, a recent study has
demonstrated that platelet-derived growth factor induces TF
expression in human SMCs by an ERK pathwaydependent mechanism and, in
part, by Ras and PKC pathwaydependent mechanisms.12 It
is possible that bFGF could act in the same manner on TF expression in
circulating monocytes and in cells of the vascular wall.
In bFGF-treated rabbits, stimulation of monocytes by endotoxin in vitro
failed to induce a further increase of TF activity. Endotoxin is known
to induce TF expression by a PKC pathwaydependent
mechanism22 leading to AP-1 and nuclear factor-
B
transcription factor activation. Because this pathway is also
stimulated by growth factors and serum,23 it seems
possible that the preliminary effect of bFGF has saturated the
endotoxin activation pathway at the second messenger or at the
transcription factor level. Interestingly, Pendurthi et
al24 recently demonstrated that bFGF pretreatment of
endothelial cells prevents TF expression in response to
phorbol myristate acetate. Taken together, these results
indicate that the cellular effects of bFGF are different
according to the stimulation context.
Effect of Hypercholesterolemia
In hypercholesterolemic rabbits, TF staining was
observed in the media (not at all in the intima), indicating that TF
induction was present in this group. By contrast, in circulating
monocytes, TF response was lower than in the control group. This
confirms data reported by our group16 and
others.25 The reasons for this lower TF activation are
unclear. The effects of lipids on monocyte TF response are complex.
Monocyte TF response is decreased by LDL and
lysophosphatidylcholine26 27 but is increased by oxidized
LDL and free cholesterol.28 29 Moreover,
hypercholesterolemia induces monocyte adhesion,
penetration, and transformation into foam cells in the vascular wall.
Because we observed that circulating monocytes decreased in
hypercholesterolemic rabbits, it seems possible that
these monocytes in hypercholesterolemic rabbits are a
subset of the total monocyte population with impaired capacity to
respond to stimulation.
Effect of bFGF Administration in Hypercholesterolemic
Rabbits
In hypercholesterolemic rabbits, bFGF induced an
increase of TF expression in circulating monocytes and in the vessel
wall. In unstimulated monocytes, bFGF restored TF activity identically
in the hypercholesterolemic and the control groups.
However, TF induction in unstimulated monocytes by bFGF remained less
important than in normocholesterolemic rabbits. By
contrast, in hypercholesterolemic rabbits treated with
bFGF, endotoxin induced a striking TF response, indicating a
"priming" effect of bFGF. In the vascular wall of
hypercholesterolemic rabbits, bFGF induced TF
expression only in the media; no TF detection was observed in the
intima. The effect of bFGF was less important than in the
normocholesterolemic rabbits.
The different effects of bFGF in hypercholesterolemic and control rabbits are unclear. Hypercholesterolemia may interfere with the relationship between bFGF and its membrane receptors. Physiologically, bFGF is stored in the extracellular matrix30 and is associated with cell surface membranes31 by heparan sulfate proteoglycan (HSPG) binding. Moreover, HSPGs are necessary for the bFGF binding to its receptors to exert both autocrine and paracrine effects.32 But atherosclerosis involves a decrease of HSPG.33 Thus, hypercholesterolemia could decrease the affinity and/or the binding of bFGF to its receptors. Conversely, hypercholesterolemia may modulate the expression of the different bFGF receptors.19 34 It is noteworthy that in this model, hypercholesterolemia did not impair the monocyte TF response to endotoxin.
Possible Consequences of Induction of TF by bFGF
TF is known to be a potent activator of blood
coagulation. The exposure of TF to blood and factor VII and its
activation could lead to thrombosis mediated by bFGF treatment, shown
particularly in the normal animals. Although this is theoretically
possible, we found no evidence of coagulation factor consumption or
decreased platelet count 3 days after the last bFGF injection.
Moreover, no evidence of arterial thrombosis was
observed.
Several previous studies indicate that TF is probably an important angiogenesis factor through mechanisms that have been discussed. TF regulates angiogenic properties of tumor cells in mice by increasing angiogenic factor (vascular endothelial growth factor, VEGF) and decreasing antiangiogenic factor (thrombospondin) production.7 TF is also involved in embryogenesis8 9 and promotes metastasis of melanoma cells.35 It was recently demonstrated that the TFfactor VIIa complex can also induce VEGF production.36 A cholesterol-rich diet in rabbits induces partial denudation of the endothelium and loss of functional properties. In our model, it was previously demonstrated that bFGF induces rapid reendothelialization, restores endothelial functions, and plays a role in angiogenesis.5 Little is known about the interaction of bFGF and TF in these processes. Importantly, it has been demonstrated in another model of vascular balloon injury that vascular cells can release both endogenous bFGF37 and TF.38
Study Limitations
The extrapolation of data from any animal model to humans requires
caution. A diet supplemented with 2% cholesterol results
in very high cholesterol levels and is associated with a
rapid progression of the vascular lesions. This
hypercholesterolemic rabbit model has been demonstrated
to have features in common with humans, and its reproducibility has
been demonstrated in several previous studies.5 39 This
model is useful to understand the processes occurring in humans and to
anticipate the effects of bFGF on human blood cells and arteries.
Administration of angiogenic growth factors is an interesting new therapeutic approach for ischemic heart disease. Administration of bFGF improves collateral vessel formation in various models of ischemia.40 41 Our results indicate that the particular dose regimen of bFGF used in this study induces TF expression in circulating monocytes and in the vascular wall in normal and in hypercholesterolemic rabbits. TF can increase the thrombogenicity of the vascular wall, and this effect can have important clinical implications. However, no evidence of consumption coagulopathy was found. Conversely, TF is probably an important angiogenic factor and can supplement the beneficial effects of angiogenic growth factor therapy. Further studies are necessary to understand the role of bFGF-induced TF expression and the therapeutic effect of bFGF.
| Acknowledgments |
|---|
Received July 19, 1999; revision received November 3, 1999; accepted November 15, 1999.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
P. Cirillo, G. Cali, P. Golino, P. Calabro, L. Forte, S. D. Rosa, M. Pacileo, M. Ragni, F. Scopacasa, L. Nitsch, et al. Tissue Factor Binding of Activated Factor VII Triggers Smooth Muscle Cell Proliferation via Extracellular Signal-Regulated Kinase Activation Circulation, June 15, 2004; 109(23): 2911 - 2916. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Tedgui and Z. Mallat Anti-Inflammatory Mechanisms in the Vascular Wall Circ. Res., May 11, 2001; 88(9): 877 - 887. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Circulation Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2000 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |