Articles |
From the Hanson Centre for Cancer Research, IMVS, Adelaide, South Australia.
Correspondence to Jennifer R. Gamble, Hanson Centre for Cancer Research, IMVS, PO Box 14 Rundle Mall, Adelaide, 5000, South Australia.
| Abstract |
|---|
|
|
|---|
(TNF-
) induces, in a time- and dose-dependent fashion, a significant
increase in its expression. Interleukin (IL)-4, IL-1, and to a lesser
extent interferon gamma have similar effects. TNF-
stimulated human
VSMCs demonstrate increased binding of T lymphocytes that is totally
VCAM-1 mediated. The cytokine transforming growth factorß (TGF-ß)
at 2.0 ng/mL inhibited basal VCAM-1 expression by 84±8% and the
induction by TNF-
by between 56±16% and 77±15% depending on the
dose of TNF. Furthermore, coculture on opposing sides of a
polycarbonate filter of human VSMCs with human umbilical vein
endothelial cells also inhibited the induction of VCAM-1 by 47±6%. As
active TGF-ß is produced upon the coculture of VSMCs and endothelial
cells, we suggest that the close physical proximity of these cells in
vivo is responsible for the lack of expression of VCAM-1 on VSMCs and
that the interruption of this contact in atheroma is an important
pathogenic event. As VCAM-1 not only serves as an adhesion molecule but
also as a costimulator of immune cells, its expression may be crucial
in the propagation of vascular lesions.
Key Words: atherogenesis T cell adhesion tumor necrosis factor interleukin-4
| Introduction |
|---|
|
|
|---|
In certain disease states such as atherogenesis or inflammation, the media is invaded by blood cells such as neutrophils, monocytes, and T lymphocytes. It is possible that these blood cells interact directly with medial SMCs by making cell-cell contacts. A consequence of these cellular interactions may be the elaboration of soluble products, eg, cytokines, which can have a profound influence on the pathogenesis of disease. To better understand the adhesion molecules involved in the interaction of SMCs with the T lymphocytes and monocytes that invade the media of blood vessels in atherogenesis, we investigated the expression of vascular cell adhesion molecule1 (VCAM-1) on SMCs.
VCAM-1 was originally described as an inducible molecule on ECs that
mediated the adhesion of lymphocytes, monocytes, and eosinophils via
the
4ß1 integrin on these blood cells.4 5 6 7 8 The special
significance of VCAM-1 in mediating the adhesion of blood cells to
endothelium is derived from its specificity for a subset of leukocytes,
its capacity to provide an integrin-mediated adhesion pathway not
involving the ß2 integrins,5 9 and its early expression
at sites where there is an accumulation of monocytes in fatty
steaks.10 11 Subsequently, the cellular expression of
VCAM-1 has broadened to include follicular dendritic cells in lymph
nodes,12 skeletal muscle cells,13 and SMCs of
atheromatous lesions.14 Indeed, VCAM-1 protein and message
have been shown to be inducible on rabbit15 and
human16 aortic SMCs, respectively.
We show here that cultured human SMCs derived from saphenous veins
express low levels of VCAM-1 in cell culture and can be induced to
express VCAM-1 by the inflammatory cytokines tumor necrosis factor
(TNF-
), interleukin (IL)-1ß, IL-4, and interferon gamma (IFN-
)
and that VCAM-1 on SMCs serves as an adhesion molecule for lymphocytes.
Furthermore, we show that coculture of human umbilical vein ECs
(HUVECs) with SMCs suppresses the expression of VCAM-1, a suppression
that is reproduced by the cytokine transforming growth factorß
(TGF-ß). As active TGF-ß is made upon SMC-EC
contact2 3 and is present in blood
vessels,17 our findings suggest that in normal blood
vessels VCAM-1 expression on SMCs is chronically depressed by their
proximity to ECs and that this depression may be mediated by TGF-ß or
another unidentified factor.
| Methods |
|---|
|
|
|---|
-SMC actin (Sigma). For the investigation of
expression of cell surface molecules, vascular SMCs (VSMCs) were plated
at 5x104 cells/100 µL into fibronectin
(Boehringer Mannheim)-coated, 24-well trays (NUNC) and grown
overnight. TGF-ß, if used, was incubated for 20 hours prior to the
addition of the other cytokines.
EC Preparation
ECs were prepared from human umbilical veins by a modified
method of Wall et al.19 The medium for growth was M199
with Earle's salts containing 20 mmol/L HEPES, 20% fetal calf serum
(GIBCO BRL), 2 mmol/L glutathione, nonessential amino acids, sodium
pyruvate, and sodium bicarbonate (Cytosystems). Cells were used at the
first passage without the addition of growth factor or heparin.
Monoclonal Antibodies
AntiVCAM-1 (51-10C9) and antiE-selectin (49-1B11) were
IgG1 monoclonal antibodies raised by the authors that
reacted specifically with Chinese hamster ovary cells transfected with
VCAM and E-selectin cDNA, respectively. 28F11 is an isotype-matched
monoclonal antibody that does not bind to unactivated or
cytokine-activated VSMCs or HUVECs. Monoclonal antiTGF-ß, a gift of
Genentech Inc, neutralizes TGF-ß1, TGF-ß2, and TGF-ß3.
Cytokines
TNF-
(2x107 U/mg; lot No. 3901DAX), IFN-
(2x107 U/mg; lot No. BN9327AX), and TGF-ß (lot No.
W9806AX) were the generous gifts of Genentech Inc. IL-1ß
(108 thymocyte mitogenic units per milligram) and IL-4 were
the generous gifts of Immunex Inc.
Coculture Experiments
For the coculture of HUVECs and VSMCs, Millipore Transwells
(Millipore) of 0.4-µm pore size and 12-mm diameter were used. Wells
were first coated with 100 µg/mL fibronectin for 2 hours and allowed
to dry overnight. To the inverted lower surface 4x104
VSMCs in 100 µL were added and allowed to adhere for 4 hours. The
chamber was then inverted into media present in the Transwell
container, and 3x104 HUVECs in 100 µL were added to the
upper surface. The cells were cocultured for 36 hours prior to
stimulation with cytokines.
Flow Cytometry
The appropriate antibody was added to cells in situ for 30
minutes at room temperature. The cells were then washed, and
fluorescein isothiocyanatecoupled anti-mouse immunoglobulin (Silenus)
was added for a further 30 minutes at room temperature. The cells were
removed by trypsin, washed once, and fixed with fluorescence-activated
cell sorter fix (paraformaldehyde, glucose, azide, and
phosphate-buffered saline, pH 7.5) prior to analysis on an Epics
Profile IV. The fluorescence was quantified as the mean channel
x10-2 of 10 000 cells counted. The shift in fluorescence
was uniform, and the median channel gave identical trends in the
results. Background fluorescence refers to the mean channel
fluorescence of cells stained with control antibody and fluorescein
isothiocyanateantimouse immunoglobulin.
Cell Adhesion Experiments
VSMCs were plated at 1.5x104/well in 96-well
microtiter plates (NUNC) and grown overnight. T cells from peripheral
blood were purified by Ficoll-Hypaque sedimentation followed by nylon
wool separation.20 The cells were labeled with
51Cr,21 and 5x105 cells were
added to the VSMC monolayers for 45 minutes at 37°C. The nonadherent
cells were removed by washing, and the adherent cells were extracted by
NH4OH. The percentage of adherent cells was calculated from
the total number of cells added.
Statistics
The two-tailed t test was used to obtain probability
values for statistical significance.
| Results |
|---|
|
|
|---|
Increases VCAM-1 Expression in Human SMCs
for 4 hours, and the expression of VCAM-1 was determined by
flow cytometry. There was a small, statistically significant expression
of VCAM-1 in the absence of TNF-
(Fig 1
(1
to 100 U/mL) caused a dose-dependent increase in VCAM-1 expression.
Although there were differences in the absolute amount of VCAM-1
induced between SMC lines, the same trend was seen in each isolate
(data not shown).
|
Time Course of TNF-
Stimulated VCAM-1 Expression in Human
SMCs
A time course of VCAM-1 expression was performed in four different
SMC lines by adding 100 U/mL TNF to aliquots of each cell line at
various times before simultaneous harvest. The results (Fig 2
) show peak expression at 5 hours after addition of
TNF-
, with waning after 18 hours but still significant levels at 48
hours.
|
IL-4 Increases VCAM-1 Expression on Human SMCs
Human SMCs were incubated for 4 to 5 hours with a range of doses
of IL-4. The results (Fig 3
) show the dose-dependent
increase in VCAM-1 expression. The addition of 10 U/mL TNF, which by
itself had a moderate effect on VCAM-1 expression, was at least
additive with IL-4. We found that IFN-
and IL-1 also increased
VCAM-1 expression on SMCs (Table 1
) and that the effects
of TNF and IFN-
and of IL-4 and IFN-
were also synergistic (data
not shown). A combination of cytokines may be much more effective in
stimulating strong VCAM-1 expression than one cytokine alone.
|
|
TGF-ß Inhibits Basal and TNF-
Stimulated VCAM-1 Expression in
Human SMCs
TGF-ß inhibits E-selectin expression in HUVECs.21
Human SMCs were incubated with 2 ng/mL TGF-ß for 24 hours before
basal VCAM-1 was measured. As shown in Fig 4
, a summary
of six separate experiments, TGF-ß diminished basal VCAM-1 expression
to near background levels (by a mean of 84±8%). The effect of 2 ng/mL
TGF-ß on VCAM-1 responses to a series of doses of TNF is also shown
in Fig 4
. At each dose there was a significant inhibition of VCAM-1
expression that ranged from a mean decrease of 77±15%, 71±17%,
70±6%, and 56±16% at TNF doses of 0.1, 1, 10, and 100 U/mL,
respectively. The effect of a range of doses of TGF-ß on 100 U/mL
TNFstimulated VCAM-1 (Fig 4
, right) demonstrates a dose-dependent
increase of inhibition.
|
ECs Inhibit VCAM-1 Expression on Human SMCs
Human SMCs were grown on the bottom of Transwell inserts with
0.4-µm pores. HUVECs were plated in the top half of some of the
inserts 24 hours before the addition of TNF-
. VCAM-1 expression was
determined on the SMC population 5 hours after the addition of TNF-
,
and as a control the effect of TGF-ß was also examined on SMCs
cultured without ECs. E-selectin expression was also measured on the
SMC and EC populations. The results of a single typical experiment are
given in Table 2
. No E-selectin expression was seen in
the SMC population, although it was highly expressed in the EC
population. These results suggested that EC contamination was not an
issue and that we were measuring SMC-associated VCAM-1 and not
EC-associated VCAM-1. In SMC-EC coculture a decrease in the basal level
of E-selectin on these primary ECs was seen, although the TNF-induced
E-selectin expression was not altered. This regulation of E-selectin
expression by coculture is under further investigation.
|
The experiment in Table 2
shows that both the basal and TNF-induced
VCAM-1 expression on SMCs was inhibited by coculture with ECs. The
pooled results of three experiments (Fig 5
) show that
VCAM-1 expression on VSMCs was reduced by 47%, a figure comparable to
the 63% reduction by TGF-ß in this series of experiments and to the
figures reported in Fig 4
. HUVECs cultured in the same dish but not on
opposite sides of the filter with VSMCs had no effect on VCAM-1
expression (data not shown).
|
TNF-
or IL-4Stimulated SMCs Support the Adhesion of T
Lymphocytes
To show the functional nature of induced VCAM-1 expression
on SMCs, the adhesion of T lymphocytes, a cell type known to adhere to
VCAM-1, was examined. Treatment of SMCs with TNF-
or IL-4 increased
adhesion (Fig 6
, left). This adhesion was inhibited by
antibody to VCAM-1 (but not to E-selectin; data not shown) and, in the
case of TNF-
, also by TGF-ß (Fig 6
, right).
|
| Discussion |
|---|
|
|
|---|
and IL-1) or T lymphocytes (IL-4 and IFN-
). These
findings are in partial agreement with Li et al15 and
Couffinhal et al.16 Li et al15 found that
IFN-
and lipopolysaccharide increased VCAM-1 message and protein in
rabbit aortic SMCs and IL-4 and IFN-
in human aortic SMCs. These
investigators failed to detect upregulation by either IL-1 or TNF-
.
In contrast, Couffinhal et al16 show a strong effect of
TNF-
and a weak one of IFN-
on VCAM-1 expression on human aortic
SMCs. No synergy was observed in these experiments, and IL-1 was found
to be without effect. Such differences may be a result of SMC source,
extraction procedures, or cell-culture conditions.
In our experience the level of expression of VCAM-1 on SMCs induced by
the cytokines TNF and IL-4 was highly significant and was reflected by
an increased adhesion of T lymphocytes (Fig 6
) that was totally VCAM-1
dependent. The role of VCAM-1 on VSMCs may extend beyond that of being
a passive adhesion structure for cells expressing
4ß1 or
4ß7
integrins.5 9 There exists evidence that VCAM-1 also
serves as a costimulatory molecule of resting or short-termstimulated
T cells.22 23 24 25 Interestingly, VCAM-1 induces apoptosis in a
proportion of long-termstimulated T cells.26 Thus,
VCAM-1 appears to have a key role in regulating T cell responses, and
as T cells from atheromatous lesions express high levels of ß1
integrins,27 it is likely that this pathway operates in
the microenvironment of atheromata. The cytokine profile of
atheromatous lesions may thus be critical in maintaining the
inflammatory-type response. The role of IL-4 is notable in this regard
as it appears to act as a switching cytokine in the expression of
adhesion molecules. On HUVECs it inhibits E-selectin
expression28 but enhances, as it does on VSMCs, VCAM-1
expression. Interestingly, IL-4 inhibits monocyte adhesion to HUVECs by
mechanisms that remain to be defined.29 These findings
place special relevance on finding the TH2 type of T cells, a chief
source of IL-4, in atheromatous lesions. However, to our knowledge the
T cells in atheromatous lesions have not been characterized beyond
expressing an activated phenotype.
The synergistic effect of TNF and IL-4 on VCAM-1 expression (Fig 3
)
suggests that the mixture of cytokines found in lesions such as
atheroma may be more inducive to VCAM-1 expression than individual
cytokines by themselves. TNF has been demonstrated by
immunocytochemical techniques to be present in atheromatous
lesions,30 and IL-4 is a product of a subset of the
activated T cells that form approximately 30% of the cellular
infiltrate of atheromatous lesions.27 31 Thus, the type of
synergy demonstrated in Fig 3
may well take place in vivo.
Interestingly, O'Brien et al14 have demonstrated VCAM-1
expression on intimal SMCs only in areas of inflammatory cell
infiltration, suggesting the possibility that factors released by these
inflammatory cells result in the induction of VCAM-1 expression. We
would postulate that on the basis of the phenotypes of T cells and
macrophages in these lesions, TNF and IL-1 are likely candidates. The
local production of cytokines would also explain the finding of SMC
activation in isolated, specific regions and not throughout the total
neointima.
The second chief finding is the strong suppression of VCAM-1 expression
that is exerted by TGF-ß on both basal and TNF-stimulated VCAM-1
expression (Fig 4
). TGF-ß inhibits endothelial E-selectin expression,
but it did not significantly alter the expression of VCAM-1 on
ECs.22 The effect of TGF-ß on VSMCs differed from HUVECs
in other aspects as well. On VSMCs the dose-response curve was linear
(Fig 4
) compared with the inverted U-shaped curve seen on
HUVECs.21 This difference cannot be explained by but may
be related to the types of receptors expressed on the two cell types or
intracellular signaling events that may follow ligand binding. The
inhibition of TGF-ß was more complete on VSMCs and operated almost
equally at all doses of TNF, whereas the inhibition seen on HUVECs was
mainly at low doses of TNF.
The third important observation presented here is that ECs
cocultured with VSMCs also inhibit VCAM-1 expression (Fig 5
). The
nature of this inhibitory effect upon coculture has yet to be defined.
One possibility is that it is mediated through TGF-ß. Active TGF-ß
is made when ECs are cocultured and in close association with another
cell type, most relevantly VSMCs.2 3 However, in our
system, the antiTGF-ß antibodies (which are known to neutralize
TGF-ß1, TGF-ß2, and TGF-ß3) that neutralized the effect of
exogenously added TGF-ß did not reproducibly neutralize the
inhibitory effect on SMCVCAM-1 expression by EC-SMC coculture (data
not shown). It is possible that the exogenously added antibody has a
variable access to the sites of endogenous production of
TGF-ß in a Transwell system. Addition of aprotinin and mannose
6-phosphate, both of which inhibit the conversion of latent to active
TGF-ß,3 32 failed to reverse the inhibitory effect of
EC-SMC coculture on VCAM-1 expression. An alternate possibility for the
inhibitory effect is that a factor other than TGF-ß may be produced
upon EC-SMC coculture. As for TGF-ß1, this factor is nondiffusible
and requires close cell growth, since ECs grown in the same well but
not on opposing sides of a filter did not regulate VCAM-1 expression on
SMCs. The nature of this factor is currently being investigated.
A central role of TGF-ß in the proper functioning of the vasculature
is supported by a number of observations. First, TGF-ß1null
mice33 34 show spontaneous multifocal inflammatory
disease, widely disseminated leukocyte infiltration, and excess
production of the proinflammatory cytokines TNF, IFN-
, and
macrophage inflammatory protein1
. Our previous
observations20 32 35 suggest that disregulated expression
of adhesion molecules on the endothelium may account for some of these
effects. Second, the high-risk factors for atherosclerosis,
lipoprotein(a) and plasminogen activator inhibitor1, block the
activation of latent TGF-ß by competitively inhibiting tissue
plasminogen activator, an enzyme essential in the conversion of latent
to active TGF-ß.36 37 TGF-ß inhibits the migration and
proliferation of VSMCs in cell culture36 37 and, as we
show here, the expression of VCAM-1. Interestingly, at sites of apo(a)
accumulation (the active moiety of lipoprotein[a]), VSMCs display an
activated phenotype, and there is a high propensity for the development
of vascular lesions at these sites.14 38 Third, patients
with advanced atherosclerosis show a significantly lower concentration
of active TGF-ß in their serum than do normal control
subjects.39 The correlation of active TGF-ß with
atherosclerosis is stronger than any of the known risk factors.
Our data also suggest the importance of EC-SMC interaction in vascular function. A physical or functional separation of these cells, such as may happen as a result of the thickening of the basal lamina40 during the early events in models of atheroma, may result in profound effects on the function of SMCs and ECs. As we have shown, the adhesiveness of VSMCs through regulation of VCAM-1 expression is controlled by ECs. Furthermore, the migration, proliferation, and production of plasminogen activator inhibitor1 by ECs is inhibited by VSMCs.2 3 41 At least one of the controlling factors produced as a result of EC-SMC apposition has been identified as TGF-ß.2 3
| Acknowledgments |
|---|
Received August 28, 1994; accepted March 21, 1995.
| References |
|---|
|
|
|---|
. J Immunol. 1990;145:865-872. [Abstract]
This article has been cited by other articles:
![]() |
H. Loppnow, K. Werdan, and M. Buerke Invited review: Vascular cells contribute to atherosclerosis by cytokine- and innate-immunity-related inflammatory mechanisms Innate Immunity, April 1, 2008; 14(2): 63 - 87. [Abstract] [PDF] |
||||
![]() |
E. P. Moiseeva Adhesion receptors of vascular smooth muscle cells and their functions Cardiovasc Res, December 1, 2001; 52(3): 372 - 386. [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] |
||||
![]() |
M. Braun, P. Pietsch, K. Schror, G. Baumann, and S. B. Felix Cellular adhesion molecules on vascular smooth muscle cells Cardiovasc Res, February 1, 1999; 41(2): 395 - 401. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Braun, P. Pietsch, A. Zepp, K. Schror, G. Baumann, and S. B. Felix Regulation of Tumor Necrosis Factor-{alpha} and Interleukin-1ß Induced Adhesion Molecule Expression in Human Vascular Smooth Muscle Cells by cAMP Arterioscler. Thromb. Vasc. Biol., November 1, 1997; 17(11): 2568 - 2575. [Abstract] [Full Text] |
||||
![]() |
F. Cambien, S. Ricard, A. Troesch, C. Mallet, L. Generenaz, A. Evans, D. Arveiler, G. Luc, J.-B. Ruidavets, and O. Poirier Polymorphisms of the Transforming Growth Factor-ß1 Gene in Relation to Myocardial Infarction and Blood Pressure: The Etude Cas-Temoin de l'Infarctus du Myocarde (ECTIM) Study Hypertension, November 1, 1996; 28(5): 881 - 887. [Abstract] [Full Text] |
||||
![]() |
M. W. Feinberg, M. K. Jain, F. Werner, N. E. S. Sibinga, P. Wiesel, H. Wang, J. N. Topper, M. A. Perrella, and M.-E. Lee Transforming Growth Factor-beta 1 Inhibits Cytokine-mediated Induction of Human Metalloelastase in Macrophages J. Biol. Chem., August 11, 2000; 275(33): 25766 - 25773. [Abstract] [Full Text] [PDF] |
||||
![]() |
Z. Mallat, A. Gojova, C. Marchiol-Fournigault, B. Esposito, C. Kamate, R. Merval, D. Fradelizi, and A. Tedgui Inhibition of Transforming Growth Factor-{beta} Signaling Accelerates Atherosclerosis and Induces an Unstable Plaque Phenotype in Mice Circ. Res., November 9, 2001; 89(10): 930 - 934. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
ATVB Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1995 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |