Download Human Vascular Smooth Muscle Cells Lack Essential Costimulatory

Human Vascular Smooth Muscle Cells Lack Essential
Costimulatory Molecules to Activate Allogeneic Memory
T Cells
Pei Zhang, Thomas D. Manes, Jordan S. Pober, George Tellides
Downloaded from http://atvb.ahajournals.org/ by guest on May 12, 2017
Objective—The arterial media, populated by vascular smooth muscle cells (VSMC), is an immunoprivileged compartment
and, in contrast to the intima or adventitia containing endothelial cells, is generally spared by inflammatory processes,
such as arteriosclerosis. To determine mechanisms of medial immunoprivilege, we investigated the ability of human
VSMC versus endothelial cells to activate allogeneic T cells in vitro.
Methods and Results—Unlike cultured endothelial cells, cultured VSMC do not activate allogeneic memory CD4 or CD8
T cells and fail to effectively support T-cell proliferation to the polyclonal activator, phytohemagglutinin, consistent
with a defect in costimulation function. Although many costimulators are comparably expressed on both cell types,
endothelial cells but not VSMC basally express OX40 ligand and upregulate inducible costimulator ligand in response
to proinflammatory cytokines. OX40 ligand-transduced, but not control- or inducible costimulator ligand-transduced,
VSMC acquire the capacity to stimulate allogeneic memory CD4 T cells to produce cytokines and to proliferate in the
presence of supplemental L-tryptophan. OX40 ligand overexpression, although not essential, also enhances allogeneic
memory CD8 T-cell responses to VSMC after L-tryptophan supplementation.
Conclusion—The inability of cultured VSMC to activate memory T cells results from a lack of essential costimulators,
particularly OX40 ligand, in addition to indoleamine 2,3-dioxygenase-mediated tryptophan depletion. (Arterioscler
Thromb Vasc Biol. 2010;30:1795-1801.)
Key Words: vascular smooth muscle cells 䡲 endothelial cells 䡲 T cells 䡲 costimulation 䡲 transplantation
T
he most common cause of death for cardiac transplant
patients is chronic rejection characterized by pathological remodeling of coronary arteries.1 Immunohistological
analyses of arteriosclerotic lesions in transplanted hearts
reveal an uneven distribution of infiltrating leukocyte within
the vessel wall. T cells are predominantly localized in the
intima and adventitia, whereas few reside within the medial
layer.1,2 Similarly, in atherosclerosis of native (untransplanted) hearts, large numbers of leukocytes accumulate in
the intima and adventitia, whereas the media is relatively free
from leukocytic infiltrates.3 These observations in diseased
vessels are mirrored in murine viral infection models, suggesting that the arterial media is an immunoprivileged site.4
Because vascular smooth muscle cells (VSMC) are the
principal and almost exclusive cell type normally found in the
media of noninflamed arteries, we have explored the hypothesis that VSMC are responsible for establishing medial
immunoprivilege. In contrast, endothelial cells (EC), which
normally line the intima and the adventitial microvessels, are
generally accepted as playing prominent proimmunogenic
roles. T cells, resident in normal arteries or infiltrating
diseased arteries, normally recognize foreign antigens as a
complex formed by antigenic peptides bound to self allelic
forms of major histocompatibility complex (MHC) molecules
expressed on the surface of antigen-presenting cells. A large
percentage of T cells can cross-react with complexes formed
by (often self) peptides with nonself allelic forms of MHC
molecules, a phenomenon known as direct recognition of
alloantigens. After pretreatment with interferon (IFN)-␥ to
reinduce MHC class II molecules, cultured human EC present
alloantigen and activate allogeneic memory CD4 T cells.5 On
IFN-␥ stimulation, cultured human VSMC are induced to
express similar levels of MHC class II antigens as EC but fail
to activate allogeneic memory CD4 T cells.6,7 In fact, IFN␥-treated, but not untreated, VSMC can actively suppress the
proliferation of CD4 T cells induced by EC by acting across
a transwell.6 Recently, we have shown that the tryptophanmetabolizing enzyme, indoleamine 2,3-dioxygenase (IDO) is
highly induced by IFN-␥ in VSMC and that tryptophan
depletion in the microenvironment by IDO inhibits T-cell
accumulation in the media.7 Although inhibiting IDO or
supplementing cultures with L-tryptophan can abrogate the
inhibitory transacting effects of IFN-␥-treated VSMC, it is
not sufficient to allow VSMC to directly activate allogeneic
Received on: November 22, 2009; final version accepted on: May 26, 2010.
From the Departments of Surgery (P.Z., G.T.) and Immunobiology (T.D.M., J.S.P.), Interdepartmental Program in Vascular Biology and Therapeutics,
Yale University School of Medicine, New Haven, Connecticut; the Veterans Affairs Connecticut Healthcare System, West Haven, Connecticut (G.T.).
Correspondence to George Tellides, 10 Amistad St, PO Box 208089, New Haven, CT 06520. E-mail [email protected]
© 2010 American Heart Association, Inc.
Arterioscler Thromb Vasc Biol is available at http://atvb.ahajournals.org
1795
DOI: 10.1161/ATVBAHA.109.200758
1796
Arterioscler Thromb Vasc Biol
September 2010
Figure 1. VSMC do not activate T cells.
A, IFN-␥-pretreated vascular cells were
cocultured with CFSE-labeled allogeneic
memory CD4 T cells for 7 days. B, class II
transactivator-transduced vascular cells
were cocultured with CFSE-labeled allogeneic memory CD4 T cells for 7 days. C,
Untreated vascular cells were cocultured
with CFSE-labeled allogeneic memory
CD8 T cells in the presence of exogenous
IL-2 for 8 days. MHC class I or II antigen
expression (black outline vs gray fill of
immunoglobulin G control) by EC and
VSMC was analyzed prior to coculture. Proliferating CFSElow T cells were measured at
completion of coculture. Data are representative of 3 independent experiments.
Downloaded from http://atvb.ahajournals.org/ by guest on May 12, 2017
CD4 T cells, suggesting additional immunomodulatory mechanisms in these cellular interactions.7
In addition to TCR engagement with antigen-bound MHC
molecules, full activation of T cells requires costimulatory
signals.8 Unlike professional antigen-presenting cells, such as
dendritic cells, which predominantly use B7 molecules to
costimulate T cells, human EC act as semiprofessional
antigen-presenting cells and use alternative costimulatory
molecules, such as leukocyte function-associated antigen-3,
OX40 ligand (OX40L), and inducible costimulator ligand
(ICOSL). This combination of costimulatory molecules allows EC to activate allogeneic memory CD4 and CD8 T cells,
but not naïve T cells.5 In this study, we have further
investigated the interactions between human VSMC and
allogeneic memory CD4 and CD8 T cells. Our data suggest
that the inability of cultured VSMC to activate T cells results
from a combination of IDO expression and a lack of OX40L
expression.
Methods
Cells
Cells and tissues were obtained under protocols approved by the
Yale University Human Investigation Committee and the New
England Organ Bank. Human EC were isolated by collagenase
harvest from umbilical cord veins.5 Human aortic and coronary
artery VSMC were isolated by explant outgrowth.7 Peripheral blood
mononuclear cells were collected by leukapheresis from healthy
donors. CD4 or CD8 T cells were positively selected from peripheral
blood mononuclear cells using Dynal magnetic beads (Invitrogen),
and memory T cells were enriched by depleting naïve T cells and
recently activated T cells with mouse antihuman CD45RA and
human leukocyte antigen-DR antibodies (eBioscience) and antimouse immunoglobulin G magnetic beads.5
Reagents and Assays
Cytokines included recombinant human tumor necrosis factor (TNF)
(R&D Systems), IFN-␥ (Biosource International), and interleukin
(IL)-2 (National Cancer Institute). Antibodies included mouse antihuman OX40L, intercellular adhesion molecule-1 (R&D Systems),
ICOSL, programmed cell death ligand (PDL)-1, PDL-2 (eBioscience), and leukocyte function-associated antigen-3 (Abcam).
LZRSpBMNZ retroviral vectors encoding LacZ or human OX40L,
ICOSL, and class II transactivator were used. We have previously
described our techniques of cell culture, cell coculture, retrovirus
transduction, fluorescence-activated cell sorter analysis, ELISA, and
quantitative PCR.5,7 Further details are provided in the supplemental
material, available online at http://atvb.ahajournals.org.
Statistical Analysis
Comparisons between 2 groups were by t test and between more than
2 groups were by ANOVA. Statistical analyses were performed
using Prism 4 software (GraphPad). Probability values were 2-tailed
and values of ⬍0.05 were considered to indicate statistical
significance.
Results
VSMC Are Unable to Activate Allogeneic
Memory T Cells
We first confirmed previous work that EC and VSMC
pretreated with IFN-␥ differ in their abilities to activate
allogeneic memory CD4 T cells.5–7 Although IFN-␥ stimulated similar expression of MHC class II molecules in both
types of vascular cells, EC but not VSMC induced proliferation of CFSE-labeled, allogeneic memory CD4 T cells as
indicated by dye dilution after 7 days of coculture (Figure
1A). Because IFN-␥ can also induce IDO expression in
VSMC at considerably greater levels than in EC or T cells
and tryptophan depletion as a result of IDO activity can
suppress T-cell clonal expansion,7 we bypassed the IFN-␥/
IDO effect by upregulating MHC class II molecule expression through retroviral-transduced class II transactivator.
However, VSMC still failed to induce allogeneic memory
CD4 T-cell proliferation (Figure 1B). We also investigated
allogeneic CD8 T-cell responses without the need for IFN-␥
pretreatment of vascular cells. Despite similar constitutive
levels of MHC class I antigens, proliferating CD8 T cells
were only detected in EC but not VSMC cocultures (Figure
1C). Addition of neutralizing antibodies to IFN-␥ to block
possible effects of endogenous cytokine in VSMC cocultures
also did not result in T-cell clonal expansion (data not
shown). Similar results were observed with different combinations of EC, VSMC, and T cell donors, ruling out a role for
specific allogeneic differences. These data demonstrate immunologic ignorance of VSMC that may result from defects
in antigen presentation, costimulation, or stable conjugate
formation.
VSMC Do Not Provide Effective Costimulation for
Allogeneic Memory T Cells
We specifically tested for vascular cell costimulation function
by adding the polyclonal activator, phytohemagglutinin
(PHA) to cocultures to cross-link the TCR and provide
adhesive interactions (assay principle further explained in
Zhang et al
VSMC Overexpressing OX40L Activate Memory T cells
1797
cocultures. Furthermore, the modest amount of CFSE dilution
triggered in VSMC cocultures suggested a single or limited
division of T cells, whereas the much greater degree of CFSE
dilution induced in EC cocultures indicated several rounds of
clonal expansion (supplemental Figure I). Similar results
were obtained when mitosis was measured by
5-bromodeoxyuridine incorporation instead of CFSE dilution
(supplemental Figure II). Together, these results are consistent with a defect in costimulation function of VSMC
compared with EC.
VSMC Lack Basal or Inducible Expression of
OX40L and ICOSL
Downloaded from http://atvb.ahajournals.org/ by guest on May 12, 2017
Figure 2. VSMC do not provide effective costimulation to T
cells. CD4 (A) and CD8 (B) CFSE-labeled memory T cells were
cocultured with allogeneic vascular cells in the presence or
absence of PHA and/or L-tryptophan (TRP). Proliferating
CFSElow T cells were measured at day 4. Bars are means⫾SEM,
and data are representative of 3 independent experiments.
experimental design section in supplemental data) with or
without L-tryptophan supplementation to overcome the effects of IDO activity. Without addition of PHA, there was
very little T-cell proliferation from either EC or VSMC
cocultures at day 4 (Figure 2A and 2B), unlike the observable
CFSE dilution of T cells in EC cocultures after 7 to 10 days
in our standard coculture assays. However, even at this early
time point, there was robust proliferation of PHA-stimulated
memory CD4 and CD8 T cells in EC cocultures but not in
VSMC cocultures. When L-tryptophan was supplemented
daily, CFSElow T cells in VSMC cocultures increased, although they did not reach the frequency of that in EC
We examined whether VSMC lack positive costimulators or
express negative costimulators in comparison with those of
EC at baseline and after cytokine treatment (Figure 3A).
Leukocyte function-associated antigen-3 expression was constitutive and of similar intensity in both cell types. In contrast,
the basal expression of OX40L on EC was more than 5-fold
higher than that on VSMC (corrected mean fluorescence
intensity [cMFI⫽specific antibody MFI⫺control immunoglobulin G cMFI] of 3 donors was 178.0⫾49.3 versus
31.7⫾22.4, P⫽0.036). Additionally, the basal and IFN-␥inducible expression of ICOSL was higher on EC compared
with that on VSMC (cMFI of 3 donors was 49.2⫾4.4 versus
14.5⫾10.4, P⫽0.037), although these levels were relatively
low. TNF increased the expression of both OX40L and
ICOSL on EC, although not on VSMC (Figure 3B and 3C).
The failure of VSMC to respond to TNF in these assays was
selective for the costimulators, because class I MHC molecules were increased under the same conditions (data not
shown). Similar basal and IFN-␥-inducible expression of
intercellular adhesion molecule-1, an adhesion molecule contributing to immunologic synapse formation and T-cell activation, was seen on EC and VSMC. Both vascular cell types
also had low basal expression of PDL-2 and similar induction
in PDL-1 and PDL-2 expression after IFN-␥ treatment.
Although in this particular donor the basal level of PDL-1 on
VSMC was higher than on EC, this difference was not
consistent among 3 donors (cMFI was 19.5⫾7.9 on EC
versus 44.5⫾23.2 on VSMC, P⫽0.35). Furthermore, blocking antibodies to PDL-1 or PDL-2 in VSMC cocultures did
not activate allogeneic T cells (data not shown). The costimu-
Figure 3. VSMC lack OX40L and ICOSL.
A, Costimulatory molecule expression
(black line vs gray line of immunoglobulin G control) by untreated or IFN-␥treated EC and VSMC. Surface expression (B) and transcript expression (C) of
OX40L and ICOSL by untreated or TNFtreated vascular cells. Bars are
means⫾SEM, and data are representative of 3 independent experiments.
1798
Arterioscler Thromb Vasc Biol
September 2010
Downloaded from http://atvb.ahajournals.org/ by guest on May 12, 2017
Figure 4. OX40L-VSMC activate CD4 T
cells. A, IFN-␥-pretreated LacZ-,
OX40L-, or ICOSL-transduced VSMC
and LacZ-transduced EC were cocultured with allogeneic memory CD4 T
cells for 48 hours, and IL-2, IFN-␥, and
IL-5 levels were determined. *P⬍0.05 vs
T cells alone (ANOVA). B and C, Cocultures were supplemented or not with
L-tryptophan (TRP) and proliferating
CFSElow T cells were measured at day 7.
D, Alternatively, cocultures were also
treated with PHA, and T-cell proliferation
was measured at day 4. *P⬍0.001 vs T
cells alone, ⫹P⬍0.001 vs no TRP
(ANOVA). Bars are means⫾SEM from 1
experiment, and data are representative
of 3 independent experiments.
lator phenotype of vascular cells suggests that the striking
defects in basal or inducible expression of OX40L and
ICOSL by VSMC may be a factor in determining their
allogenicity.
Inducible Expression of OX40 and ICOS by T
Cells in Coculture With Vascular Cells
To determine whether the absence of OX40L and ICOSL on
VSMC could be of functional significance in our coculture
system, we investigated the expression of their respective
receptors on human memory T cells. Because both TCR and
costimulation signals can upregulate the expression of OX40
and ICOS,8 we tested the effects of vascular cells with or
without PHA. Polyclonal TCR signals alone upregulated the
expression of OX40 and particularly ICOS on memory CD4
and CD8 T cells (supplemental Figure III). In EC cocultures,
there was further induction of the costimulator receptors; the
effects were greater for OX40 than for ICOS expression and
for CD4 T cells than for CD8 T cells. In VSMC cocultures,
there were minimal changes except for OX40 induction on
CD8 T cells. These results confirmed that OX40 and ICOS
are expressed by activated memory T cells in coculture with
vascular cells and that ligation of these receptors may
contribute to T-cell activation in such cocultures.
OX40L-Transduced VSMC Induce CD4 T-Cell
Cytokine Production and Proliferation
To directly test our hypothesis that differences in OX40L
and/or ICOSL expression between EC and VSMC contribute
to the different capacities of these cell types to activate
allogeneic memory T cells, we generated VSMC transductants expressing these molecules, as well as control transductants expressing LacZ, by means of retroviral vectors. OX40L
or ICOSL surface expression on transduced VSMC was
comparable with endogenous levels on EC (supplemental
Figure IV). MHC class II molecule expression was achieved
by IFN-␥ pretreatment, because VSMC were relatively refractory to repeat transduction with retrovirus encoding class
II transactivator.
We then examined the effects of IFN-␥-pretreated VSMC
transductants on cytokine production by CD4 T cells at 48
hours of coculture (Figure 4A). CD4 T cells alone did not
secrete cytokines, whereas CD4 T cells from EC cocultures
did. Interestingly, significant levels of IL-2, IFN-␥, and IL-5
were detected from OX40L-VSMC cocultures. In contrast,
there was no significant induction of cytokine secretion from
LacZ- or ICOSL-VSMC cocultures. IL-10 and IL-17 levels
were very low (below the minimum ELISA standards) and
inconsistent between VSMC cocultures.
Despite the effect on IL-2 production, OX40L overexpression on VSMC by itself was not sufficient to stimulate
Zhang et al
VSMC Overexpressing OX40L Activate Memory T cells
1799
Downloaded from http://atvb.ahajournals.org/ by guest on May 12, 2017
Figure 5. OX40L-VSMC augment CD8
T-cell activation. A, LacZ-, OX40L-, or
ICOSL-transduced VSMC and LacZtransduced EC were cocultured with
allogeneic memory CD8 T cells and PHA
at various concentrations for 48 hours,
and IL-2, IFN-␥, and TNF levels were
determined. *P⬍0.05 vs LacZ-VSMC
(ANOVA). B and C, Cocultures with
exogenous IL-2 were supplemented or
not with L-tryptophan (TRP), and proliferating CFSElow T cells were measured at
day 8. D, Alternatively, cocultures were
performed without exogenous IL-2, and
proliferation was assessed later at day
10. *P⬍0.001 vs T cells alone,
⫹P⬍0.001 vs no TRP (ANOVA). Bars are
means⫾SEM from 1 experiment, and
data are representative of 3 independent
experiments.
allogeneic memory CD4 T-cell proliferation at day 7 of
coculture (Figure 4B and 4C). However, daily supplementation with L-tryptophan, which had little effect on T-cell
proliferation from EC cocultures, resulted in a large increase
in CFSElow cells from OX40L-VSMC cocultures and smaller
increases from LacZ- and ICOSL-VSMC cocultures. The
differences in CD4 T-cell proliferation between OX40LVSMC versus LacZ-VSMC cocultures reached statistical
significance (P⬍0.001). Similar results were obtained when
independent cocultures were performed in the presence of
PHA as a more specific test of costimulator function (Figure
4D). Additionally, proliferating CD4 T cells from OX40LVSMC cocultures supplemented with L-tryptophan also expressed the IL-2 receptor activation marker, CD25 to a
similar extent as from LacZ-EC cocultures (supplemental
Figure V). As a further specificity control, the proliferative
effect of OX40L-VSMC transductants on CD4 T cells was
inhibited by neutralizing antibody to OX40L (supplemental
Figure VI). In sum, our data demonstrate that overexpression
of OX40L, but not ICOSL, provides effective costimulation
for VSMC to activate allogeneic memory CD4 T cells when
the growth suppressive effects of IDO-mediated tryptophan
depletion are prevented.
OX40L-Transduced VSMC Augment CD8 T-Cell
Cytokine Production and Proliferation
We next examined the capacity of VSMC overexpressing
OX40L or ICOSL to activate allogeneic memory CD8 T cells
in coculture. At 48 hours, very little IL-2, IFN-␥, or TNF
were produced by CD8 T cells from VSMC transductant
cocultures (Figure 5A). Because memory CD8 T cells may
produce less cytokines than memory CD4 T cells in response
to alloantigen, we used PHA to engage the TCR on a larger
population of T cells to enhance cytokine production. With
increasing doses of PHA, greater cytokine secretion by CD8
T cells was found in all 3 VSMC cocultures, although
OX40L-VSMC consistently induced higher levels of cytokines and ICOSL-VSMC had little effect on IL-2 levels
compared with LacZ-VSMC.
We also assessed CD8 T-cell proliferation at day 8 of
coculture with transduced vascular cells in the presence of a
suboptimal concentration of exogenous IL-2 to augment the
magnitude of the proliferative responses. Without addition of
L-tryptophan, only LacZ-EC induced CD8 T-cell proliferation
(Figure 5B and 5C). With daily supplementation of
L-tryptophan, there was an increased frequency of proliferating T cells from all 3 VSMC cocultures. Similar results were
obtained when the cocultures were performed without exogenous IL-2 (to avoid potential masking of transduced costimulator molecule function), although with fewer CFSElow
cells even when assessed at a later time point of day 10
(Figure 5D). The small differences between CD8 T-cell
proliferation induced by OX40L-VSMC versus LacZ-VSMC
transductants were enhanced in the absence of IL-2 and
reached statistical significance (P⬍0.05). Overall, our experiments demonstrate that OX40L overexpression, although not
1800
Arterioscler Thromb Vasc Biol
September 2010
essential, confers increased costimulatory function to VSMC
to activate allogeneic memory CD8 T cells.
Discussion
Downloaded from http://atvb.ahajournals.org/ by guest on May 12, 2017
In this study, we find that: (1) the absence of essential
costimulator function combined with the suppressive effects
of IDO-mediated tryptophan depletion are the basis as to why
VSMC, unlike EC, fail to activate allogeneic memory CD4
and CD8 T cells; (2) in comparison with human EC, human
VSMC lack basal or cytokine-inducible expression of OX40L
and ICOSL; and (3) retroviral-mediated overexpression of
OX40L induces VSMC immunogenicity.
Previous studies have reported that human VSMC, unlike EC,
do not activate allogeneic CD4 T cells in coculture,5–7,9 –11
although once T cells are activated by EC, they are capable of
responding to VSMC from the same donor.6 The latter
observation underscores the less stringent costimulation requirements for recently activated T cells compared with
resting T cells, and establishes that T cells can recognize
antigen presented by VSMC. In our earlier work, we found
that VSMC inhibit memory CD4 T-cell responses to allogeneic EC across a transwell due to IFN-␥-inducible IDO
expression and tryptophan depletion.7 Our present work does
not provide evidence for additional IFN-␥-inducible immunomodulatory properties of VSMC. Instead, we demonstrate
that impaired costimulation function contributes to immunologic ignorance of allogeneic VSMC by T cells. Previous
studies have also reported that allogeneic VSMC do not
activate CD8 T cells in coculture.11 A new finding of our
study is that tryptophan depletion by VSMC inhibits activation of memory CD8 T cells in addition to that of memory
CD4 T cells. Our conclusions regarding the expression of
costimulator molecules in human vascular cells may not
apply to rodents, as it has been reported that murine
VSMC, but not EC, activate CD4 T-cell cytokine production and proliferation.12–14
The deficiency in memory CD4 T-cell activation by human
VSMC is at least partially rescued by the overexpression of
OX40L to similar levels as basal expression on EC. IL-2,
IFN-␥, and IL-5 secretion are induced by OX40L-expressing
VSMC, and there is no evidence for obvious immune deviation in our assays. Clonal expansion of alloreactive memory
CD4 T cells in response to OX40L-VSMC occurs after
tryptophan supplementation. These findings are consistent
with the known role of OX40L-OX40 signaling in enhancing
the production of both T helper 1 and T helper 2 cytokines
under certain conditions, in promoting effector and memory
CD4 T-cell proliferation and survival, and in regulating the
recall response of memory CD4 T cells.15–17 Although not
essential, OX40L-expressing VSMC also modestly increase
cytokine production and proliferation of allogeneic memory
CD8 T cells. The role of OX40L-OX40 signaling is less well
established for costimulation of CD8 T cells but appears to
play a nonredundant role in augmenting memory generation
and recall responses by CD8 T cells in some types of immune
responses18,19 although not others.20 In line with their action
in controlling the extent of T-cell priming following antigen
recognition, the expression of OX40 and its ligand is inducible in immune cells. OX40 is not expressed on resting T
cells, but is rapidly reinduced on memory T cells following
reencounter with antigen, more transiently on CD8 than CD4
T cells.17 Similarly, there is little basal expression of OX40L
on resting professional antigen-presenting cells, eg, dendritic
cells, but is rapidly induced by inflammatory signals.21 In
contrast, OX40L has relatively high basal expression on
human EC5 and is expressed by other human cell types,
including airway smooth muscle cells.22 We have confirmed
that human VSMC from normal or diseased aortas do not
express OX40L in situ (supplemental Figure VII). The
mechanism(s) that prevents the basal or inducible expression
of OX40L on VSMC is unknown and is of interest in the basis
of medial immunoprivilege.
In contrast to our findings with OX40L, the overexpression
of ICOSL in VSMC to levels that mirror endogenous basal
expression by EC only modestly increase effector cytokine
secretion but do not augment IL-2 production or increase
memory CD4 and CD8 T-cell proliferation compared with
control LacZ-VSMC transductants. A limitation of our studies to determine the role of ICOSL is its resistance to TNF
induction in VSMC. This phenomenon is particularly notable
as TNF upregulates the expression of ICOSL more than
10-fold on EC under the same conditions. ICOSL was
originally cloned as a TNF-inducible gene.23 This study and
others have shown that TNF can induce surface ICOSL on
diverse human cell types, including fibroblasts, EC, and
skeletal muscle cells.23–25 Similar to that for OX40L, the
mechanism(s) preventing ICOSL expression by TNF in
VSMC and whether other stimuli can induce ICOSL expression in VSMC are unknown.
In the absence of OX40L and ICOSL expression, other
costimulatory molecules may have contributed to the limited
proliferation of memory CD4 and CD8 T cells induced by
nontransduced VSMC in the presence of PHA and
L-tryptophan and to the limited proliferation of memory CD8
T cells induced by LacZ-VSMC transductants after
L-tryptophan supplementation. Candidates include leukocyte
function-associated antigen-3 and intercellular adhesion
molecule-1, which are equivalently expressed on VSMC and
EC.6 We did not examine the expression of B7 molecules,
because both VSMC and EC are known to be deficient in
these CD28 ligands.6 Of relevance, memory CD4 T cells are
less dependent on costimulation by B7 molecules compared
with naïve CD4 T cells. OX40 and ICOS signaling are
important for memory T-cell activation and are causally
implicated in animal models of disease,26 but their role in
human arterial diseases is not well defined beyond associations linking genetic variants of OX40L and OX40 to
myocardial infarction.27,28
We have extended the concept of medial immunoprivilege
developed in murine viral infection models4 to describe the
well-characterized sparing of the arterial media by leukocytic
infiltrates in many inflammatory arterial diseases. A possible
barrier to T-cell trafficking through the media is the presence
of circumferential elastic fibers between every layer of
VSMC. Disruption of elastic laminae by immune cells in
advanced atherosclerosis29 may provide a mechanical explanation for the breakdown in medial immunoprivilege at this
stage of the disease. An alternative, although not mutually
Zhang et al
VSMC Overexpressing OX40L Activate Memory T cells
exclusive, biological explanation of this phenomenon is the
regulation of T-cell responses by VSMC. Although direct
measurements of tryptophan levels in tissue compartments
have not been performed in vivo, the inhibitory properties of
VSMC on T-cell activation within the media depend on
IDO-mediated effects of this normally unperfused compartment within the arterial wall.7 The absence of positive
costimulators on resident VSMC, demonstrated here, is likely
to contribute as well. In certain circumstances, medial immunoprivilege breaks down, resulting in transmural inflammation as seen in inflammatory aneurysms.30 In arterial diseases
with more attenuated immune responses, such as graft arteriosclerosis (due to concomitant immunosuppression) and
atherosclerosis (due to limited antigenicity of putative autoantigens), intact medial immunoprivilege may prevent
transmural vasculitis and full-thickness vessel wall injury and
modify the disease process.
13.
14.
15.
16.
17.
18.
Downloaded from http://atvb.ahajournals.org/ by guest on May 12, 2017
Sources of Funding
This work was supported by the National Institutes of Health Grants
PO1 HL70295 and RO1 HL051014. P.Z. was supported by a James
Hudson Brown–Alexander B. Coxe fellowship award, Yale
University.
19.
20.
Disclosures
None.
21.
References
1. Billingham ME. Cardiac transplant atherosclerosis. Transplant
Proc. 1987;19:19 –25.
2. van Loosdregt J, van Oosterhout MF, Bruggink AH, van Wichen DF, van
Kuik J, de Koning E, Baan CC, de Jonge N, Gmelig-Meyling FH, de
Weger RA. The chemokine and chemokine receptor profile of infiltrating
cells in the wall of arteries with cardiac allograft vasculopathy is
indicative of a memory T-helper 1 response. Circulation. 2006;114:
1599 –1607.
3. Watanabe M, Sangawa A, Sasaki Y, Yamashita M, Tanaka-Shintani M,
Shintaku M, Ishikawa Y. Distribution of inflammatory cells in adventitia
changed with advancing atherosclerosis of human coronary artery.
J Atheroscler Thromb. 2007;14:325–331.
4. Dal Canto AJ, Swanson PE, O’Guin AK, Speck SH, Virgin HW. IFN-␥
action in the media of the great elastic arteries, a novel immunoprivileged
site. J Clin Invest. 2001;107:R15–R22.
5. Shiao SL, McNiff JM, Pober JS. Memory T cells and their costimulators
in human allograft injury. J Immunol. 2005;175:4886 – 4896.
6. Murray AG, Libby P, Pober JS. Human vascular smooth muscle cells
poorly co-stimulate and actively inhibit allogeneic CD4⫹ T cell proliferation in vitro. J Immunol. 1995;154:151–161.
7. Cuffy MC, Silverio AM, Qin L, Wang Y, Eid R, Brandacher G, Lakkis
FG, Fuchs D, Pober JS, Tellides G. Induction of indoleamine 2,3-dioxygenase in vascular smooth muscle cells by interferon-␥ contributes to
medial immunoprivilege. J Immunol. 2007;179:5246 –5254.
8. Croft M. Co-stimulatory members of the TNFR family: keys to effective
T-cell immunity? Nat Rev Immunol. 2003;3:609 – 620.
9. Pober JS, Collins T, Gimbrone MA Jr, Libby P, Reiss CS. Inducible
expression of class II major histocompatibility complex antigens and the
immunogenicity of vascular endothelium. Transplantation. 1986;41:
141–146.
10. Salomon RN, Hughes CC, Schoen FJ, Payne DD, Pober JS, Libby P.
Human coronary transplantation-associated arteriosclerosis. Evidence for
a chronic immune reaction to activated graft endothelial cells. Am J
Pathol. 1991;138:791–798.
11. McDouall RM, Page CS, Hafizi S, Yacoub MH, Rose ML. Alloproliferation of purified CD4⫹ T cells to adult human heart endothelial cells,
and study of second-signal requirements. Immunology. 1996;89:220 –226.
12. Fabry Z, Sandor M, Gajewski TF, Herlein JA, Waldschmidt MM, Lynch
RG, Hart MN. Differential activation of Th1 and Th2 CD4⫹ cells by
22.
23.
24.
25.
26.
27.
28.
29.
30.
1801
murine brain microvessel endothelial cells and smooth muscle/pericytes.
J Immunol. 1993;151:38 – 47.
Swanson BJ, Baiu DC, Sandor M, Fabry Z, Hart MN. A small population
of vasculitogenic T cells expands and has skewed T cell receptor usage
after culture with syngeneic smooth muscle cells. J Autoimmun. 2003;
20:125–133.
Kreisel D, Krasinskas AM, Krupnick AS, Gelman AE, Balsara KR,
Popma SH, Riha M, Rosengard AM, Turka LA, Rosengard BR. Vascular
endothelium does not activate CD4⫹ direct allorecognition in graft
rejection. J Immunol. 2004;173:3027–3034.
Gramaglia I, Jember A, Pippig SD, Weinberg AD, Killeen N, Croft M.
The OX40 costimulatory receptor determines the development of CD4
memory by regulating primary clonal expansion. J Immunol. 2000;165:
3043–3050.
Salek-Ardakani S, Song J, Halteman BS, Jember AG, Akiba H, Yagita H,
Croft M. OX40 (CD134) controls memory T helper 2 cells that drive lung
inflammation. J Exp Med. 2003;198:315–324.
Croft M, So T, Duan W, Soroosh P. The significance of OX40 and
OX40L to T-cell biology and immune disease. Immunol Rev. 2009;229:
173–191.
Chen AI, McAdam AJ, Buhlmann JE, Scott S, Lupher ML Jr, Greenfield
EA, Baum PR, Fanslow WC, Calderhead DM, Freeman GJ, Sharpe AH.
Ox40-ligand has a critical costimulatory role in dendritic cell:T cell
interactions. Immunity. 1999;11:689 – 698.
Song A, Tang X, Harms KM, Croft M. OX40 and Bcl-xL promote the
persistence of CD8 T cells to recall tumor-associated antigen. J Immunol.
2005;175:3534 –3541.
Kopf M, Ruedl C, Schmitz N, Gallimore A, Lefrang K, Ecabert B,
Odermatt B, Bachmann MF. OX40-deficient mice are defective in Th cell
proliferation but are competent in generating B cell and CTL Responses
after virus infection. Immunity. 1999;11:699 –708.
Ohshima Y, Tanaka Y, Tozawa H, Takahashi Y, Maliszewski C,
Delespesse G. Expression and function of OX40 ligand on human dendritic cells. J Immunol. 1997;159:3838 –3848.
Burgess JK, Carlin S, Pack RA, Arndt GM, Au WW, Johnson PR, Black
JL, Hunt NH. Detection and characterization of OX40 ligand expression
in human airway smooth muscle cells: a possible role in asthma?
J Allergy Clin Immunol. 2004;113:683– 689.
Swallow MM, Wallin JJ, Sha WC. B7h, a novel costimulatory homolog
of B7.1 and B7.2, is induced by TNF␣. Immunity. 1999;11:423– 432.
Wiendl H, Mitsdoerffer M, Schneider D, Melms A, Lochmuller H,
Hohlfeld R, Weller M. Muscle fibres and cultured muscle cells express
the B7.1/2-related inducible co-stimulatory molecule, ICOSL: implications for the pathogenesis of inflammatory myopathies. Brain. 2003;
126:1026 –1035.
Klingenberg R, Autschbach F, Gleissner C, Giese T, Wambsganss N,
Sommer N, Richter G, Katus HA, Dengler TJ. Endothelial inducible
costimulator ligand expression is increased during human cardiac
allograft rejection and regulates endothelial cell-dependent allo-activation
of CD8⫹ T cells in vitro. Eur J Immunol. 2005;35:1712–1721.
Gotsman I, Sharpe AH, Lichtman AH. T-cell costimulation and coinhibition in atherosclerosis. Circ Res. 2008;103:1220 –1231.
Wang X, Ria M, Kelmenson PM, Eriksson P, Higgins DC, Samnegård A,
Petros C, Rollins J, Bennet AM, Wiman B, de Faire U, Wennberg C,
Olsson PG, Ishii N, Sugamura K, Hamsten A, Forsman-Semb K, Lagercrantz J, Paigen B. Positional identification of TNFSF4, encoding OX40
ligand, as a gene that influences atherosclerosis susceptibility. Nat Genet.
2005;37:365–372.
Ria M, Eriksson P, Boquist S, Ericsson CG, Hamsten A, Lagercrantz J.
Human genetic evidence that OX40 is implicated in myocardial
infarction. Biochem Biophys Res Commun. 2006;339:1001–1006.
Gräbner R, Lötzer K, Döpping S, Hildner M, Radke D, Beer M,
Spanbroek R, Lippert B, Reardon CA, Getz GS, Fu YX, Hehlgans T,
Mebius RE, van der Wall M, Kruspe D, Englert C, Lovas A, Hu D,
Randolph GJ, Weih F, Habenicht AJ. Lymphotoxin ␤ receptor signaling
promotes tertiary lymphoid organogenesis in the aorta adventitia of aged
ApoE⫺/⫺ mice. J Exp Med. 2009;206:233–248.
Tang PC, Yakimov AO, Teesdale MA, Coady MA, Dardik A, Elefteriades JA, Tellides G. Transmural inflammation by interferon-␥producing T cells correlates with outward vascular remodeling and
intimal expansion of ascending thoracic aortic aneurysms. FASEB J.
2005;19:1528 –1530.
Downloaded from http://atvb.ahajournals.org/ by guest on May 12, 2017
Human Vascular Smooth Muscle Cells Lack Essential Costimulatory Molecules to
Activate Allogeneic Memory T Cells
Pei Zhang, Thomas D. Manes, Jordan S. Pober and George Tellides
Arterioscler Thromb Vasc Biol. 2010;30:1795-1801; originally published online June 10, 2010;
doi: 10.1161/ATVBAHA.109.200758
Arteriosclerosis, Thrombosis, and Vascular Biology is published by the American Heart Association, 7272
Greenville Avenue, Dallas, TX 75231
Copyright © 2010 American Heart Association, Inc. All rights reserved.
Print ISSN: 1079-5642. Online ISSN: 1524-4636
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://atvb.ahajournals.org/content/30/9/1795
Data Supplement (unedited) at:
http://atvb.ahajournals.org/content/suppl/2010/06/10/ATVBAHA.109.200758.DC1
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Arteriosclerosis, Thrombosis, and Vascular Biology can be obtained via RightsLink, a service of the
Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for
which permission is being requested is located, click Request Permissions in the middle column of the Web
page under Services. Further information about this process is available in the Permissions and Rights
Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Arteriosclerosis, Thrombosis, and Vascular Biology is online
at:
http://atvb.ahajournals.org//subscriptions/
Supplement Material
Expanded Methods
Cell Culture
EC and VSMC were cultured in M199 medium with 20% heat-inactivated FCS (Lonza),
2 mM L-glutamine, 100 U/mL penicillin, and 100 µg/mL streptomycin (Invitrogen). EC cultures
were also supplemented with 100 µg/mL heparin (Sigma-Aldrich) and 50 µg/mL EC growth
supplement (Collaborative Biomedical Products). Cells were used between passages 2 and 6.
Co-Culture Experiments
1 x 105 EC or VSMC were plated onto 0.1% gelatin-coated wells of 24-well plates (BD
Falcon) and after several days 1 x 106 allogeneic memory CD4 or CD8 T cells labeled with 250
nM CFSE (Molecular Probes) were introduced. In experiments involving CD4 T cells, the
vascular cells were either pretreated with IFN-γ at 100 ng/mL for 72 h or were transduced with
retrovirus encoding CIITA to induce the expression of MHC class II molecules. All co-cultures
were conducted in RPMI 1640 medium supplemented with 10% FCS, glutamine, and antibiotics.
T cell proliferation was measured by FACS after counterstaining with PE-conjugated CD4 or
CD8 antibodies (BD Pharmingen) and assessing CFSE dilution or BrdU incorporation using an
APC BrdU Flow kit (BD Pharmingen) according to the manufacturer’s instructions. In certain
studies of CD8 T cell proliferation, exogenous IL-2 at 10 U/mL was added on day 0. In other
experiments, we added PHA (Sigma-Aldrich) at 1 µg/mL on day 0 and/or L-tryptophan (SigmaAldrich) at 24.5 µM daily. Of relevance, M199 medium contains 49 µM L-tryptophan and RPMI
1640 medium contains 24.5 µM L-tryptophan. In additional control studies, mouse anti-human
OX40L antibody (R&D Systems) or isotype-matched IgG were added to the vascular cell
transductants 30 min prior to co-culture with CD4 T cells at 10 µg/mL for 7 days.
1
Cloning and Retroviral-Mediated Transduction
OX40L and ICOSL cDNA were reverse-transcribed from total RNA derived from TNFtreated EC using primers for OX40L (forward, 5’-ATGGAAAGGGTCCAACCCCTGGAA-3’;
reverse, 5’-TCAAAGGACACAGAATTCACCAGG-3’) and for ICOSL (forward, 5’-ATGCGG
CTGGGCAGTCCTGGACTG-3’; reverse, 5’-TCAAACGTGGCCAGTGAGCTCTGT-3’).
OX40L or ICOSL cDNA were inserted into LZRSpBMNZ retroviral vectors (kindly provided by
Dr. G. P. Nolan, Stanford University) through the multiple cloning sites. The correct orientation
and the DNA sequence of the insertion were verified by sequencing. CIITA- or LacZ- containing
LZRSpBMNZ retroviral constructs. The retroviral constructs were transfected into the PhoenixAmpho packaging cell line by using Lipofectamine 2000 (Invitrogen). Puromycin-resistant
Phoenix-Ampho packaging cells were selected and used to condition the medium. Retrovirusconditioned medium was passed through 0.45 µm filters (Millipore), supplemented with 8
µg/mL polybrene (Sigma-Aldrich), and used to transduce vascular cell cultures.
FACS Analysis
Cultured vascular cells were analyzed after no treatment in M199 medium supplemented
with 20% FCS or after treatment with IFN-γ at 100 ng/mL for 72 h or TNF at 10 ng/mL for 72 h.
For direct labeling, we used FITC-conjugated CD4, CD8, CD25, and PE-conjugated OX40 mAb
from BD Pharmingen, as well as APC-conjugated HLA-DR, HLA-ABC, and PE-conjugated
ICOS mAb from eBioscience. For indirect labeling, cells were incubated with primary mAb as
listed under Reagents at 5 µg/mL for 30 min at 4 C followed by biotin-conjuated goat antimouse antibody (Jackson ImmunoResearch Laboratories) at 2 µg/mL for 1 h at 4 C. After two
washes, the cells were incubated with 1 µg/mL strepavidin-R PE (Molecular Probes) at 4 C
2
before analysis on a FACS Calibur (BD Biosciences). Expression was calculated as corrected
mean fluorescence intensity (cMFI = Specific Antibody MFI – Control IgG MFI).
ELISA
Supernatants were collected from co-cultures of vascular cells and T cells at 48 h and
cytokine levels were determined using Duoset ELISA kits from R&D Systems.
Quantitative RT-PCR
Total RNA was extracted from vascular cells, either untreated in M199 medium
supplemented with 20% FCS or after TNF treatment at 10 ng/mL for 72 h, using RNeasy Mini
Kits (Qiagen). Bulk RT with random hexamer primers were performed according to the
Multiscribe RT system protocol (Applied Biosystems). Real-time PCR were performed using
Taqman mastermix and predeveloped probes for ICOSL and GAPDH or a customized probe for
OX40L (Applied Biosystem). An iCycler and its system interface software (Bio-Rad
Laboratories) were used to run the samples and analyze the data. Samples were run in duplicate
and DNase-treated RNA samples processed without RT enzyme were used as negative controls.
Transcript expression levels were normalized to that of GAPDH.
Aortic Specimens
Full-thickness biopsies from the proximal thoracic aorta were obtained from 3 patients
with ascending thoracic aortic aneurysms undergoing surgical repair and from 3 subjects with
non-aneurysmal ascending thoracic aortas undergoing cardiac transplantation or organ
procurement.
3
Immunofluorescence Imaging
Frozen sections of OCT-embedded aortas were stained with rabbit anti-human OX40L
Ab (Santa Cruz Biotechnology) followed by an Alexa Fluor® 594 donkey anti-rabbit secondary
Ab (Molecular Probes). The same artery slides were then stained with either mouse anti-human
CD45-FTIC (Beckman Coulter), CD31-FITC (R&D Systems), or α-smooth muscle actin-FITC
(Sigma-Aldrich). After counterstaining with ProLong DAPI (Molecular Probes), the slides were
analyzed using an Axiovert epifluorescence microscope (Carl Zeiss Microimaging) and
OPENLAB software (Improvision).
4
Experimental Design
PHA Costimulation Assay
In the cell co-culture system, T cell activation depends upon at least three interactions
between the T cell and the APC: MHC presentation of antigen, costimulation, and stable
conjugate formation. It is possible that the MHC molecules displayed by EC and VSMC are
bound with different peptides and this could contribute to differences in allogeneic T cell
activation. It is also clear that memory T cells are largely defined by surface proteins that allow
adhesion to EC, and T cells may be less able to form stable conjugates with VSMC. Both of
these issues can be controlled by supplementing co-cultures with the plant lectin, PHA.
Specifically, PHA binds to human TCR, providing a polyclonal signal for activation that
bypasses the need for MHC engagement. PHA is multivalent and also promotes stable conjugate
formation between T cells and APC that bypasses the need for specific adhesive interactions.
PHA does not provide T cells with costimulation, and thus its addition to a co-culture allows a
direct assessment of the function of costimulator molecules expressed by an APC. For our
experiments, we used a suboptimal dose of PHA at 1 µg/mL which did not result in memory T
cell proliferation. A higher dose of 3 µg/mL PHA triggered proliferation of memory CD4 T cells
(but not that of CD8 T cells) even in the absence of additional APC (%CFSElow cells in CD4 T
cell only cultures was 9.2±3.0 in conventional medium vs. 11.8±1.2 with L-tryptophan
supplementation). Since co-cultures with PHA involve multiple T cell clones, rather than only
those alloreactive with the APC donor, the measurable proliferative response of T cells to
competent APC is accelerated by several days and CFSE dilution is assessed at day 4 of coculture rather than at day 7-8 without PHA. At the early time point of day 4, proliferation of T
cells is not detectable in EC co-cultures in the absence of PHA.
5
IL-2 Supplementation of CD8 T Cell Co-Cultures
In studies of CD8 T cell proliferation, a suboptimal concentration of IL-2 was added on
day 0 to augment the relatively low proliferative rates since many CD8 responses depend upon
IL-2 provided by concomitantly activated CD4 T cells. Preliminary experiments had established
that increasing concentrations of IL-2 progressively increased the fraction of proliferating cells
only in EC co-cultures. At the maximal IL-2 concentration of 100 U/mL, the %CFSElow cells
were 42.8% in EC co-cultures, 0.12% in VSMC co-cultures, and 1.17% in CD8 T cell alone
cultures at day 8. In our experiments we used a suboptimal concentration of 10 U/mL IL-2 added
as a single dose at day 0 of co-culture. In certain control experiments, exogenous IL-2 was not
added to avoid potential masking of transduced costimulator molecule function.
6
Supplemental Figure Legends
Supplemental Figure I. VSMC do not provide effective costimulation to T cells. (A) CD4 and
(B) CD8 CFSE-labeled memory T cells were co-cultured with allogeneic vascular cells in the
presence or absence of PHA at 1 µg/mL on day 0 and/or L-tryptophan (TRP) at 24.5 µM daily.
Proliferating T cells were measured at day 4 as % CFSElow cells (upper left quadrant). Data are
representative of three independent experiments.
Supplemental Figure II. VSMC do not provide effective costimulation to T cells. (A) CD4 and
(B) CD8 memory T cells were co-cultured with allogeneic vascular cells in the presence or
absence of PHA at 1 µg/mL on day 0 and/or L-tryptophan (TRP) at 24.5 µM daily and pulsed
with BrdU at 10 µM on day 3. Proliferating T cells were measured at day 4 as % BrdU+ cells
(right side of histogram). Data are representative of two independent experiments.
Supplemental Figure III. OX40 and ICOS expression by T cells. (A) CD4 and (B) CD8
memory T cells were co-cultured with allogeneic vascular cells in the presence or absence of
PHA and/or L-tryptophan (TRP). OX40 and ICOS expression was measured at day 4. Bars are
means±SEM. *p<0.05 vs. Medium (ANOVA).
Supplemental Figure IV. Overexpression of OX40L and ICOSL by VSMC. (A) VSMC were
transduced with retrovirus vectors encoding OX40L, ICOSL, or LacZ and the expression levels
of costimulatory molecules were compared to that of EC transduced with LacZ. Histograms of
OX40L and ICOSL expression by LacZ-EC (upper panels), OX40L-VSMC and LacZ-VSMC
(bottom left panel), and ICOSL-VSMC and LacZ-VSMC (bottom right panel) are shown (thick
black line, OX40L or ICOSL transductants; thin black line, LacZ transductants; gray fill, IgG
7
controls). (B) The expression of OX40L, ICOSL, and MHC class II molecules by LacZ-EC
(upper panels), OX40L-VSMC (bottom left and middle panels), and ICOSL-VSMC (bottom
right panel) was also analyzed after IFN-γ treatment at 100 ng/mL for 72 h. Data are
representative of three independent experiments.
Supplemental Figure V. OX40L-transduced VSMC induce CD25 expression by CD4 T cells.
IFN-γ-pretreated LacZ-, OX40L-, or ICOSL-transduced VSMC and LacZ-transduced EC were
co-cultured with allogeneic memory CD4 T cells for 7 days with or without supplemental
tryptophan (TRP). T cell proliferation was measured by CFSE dilution and CD25 expression was
also analyzed by FACS. The %CFSElowCD25high cells are shown in the upper left quadrants.
Data are from one of three independent experiments using different donors with similar results.
Supplemental Figure VI. Neutralizing OX40L antibodies inhibit OX40L-VSMC-induced CD4
T cell proliferation. (A) IFN-γ-pretreated LacZ-transduced EC, LacZ-transduced VSMC, or
OX40L-transduced VSMC were co-cultured with allogeneic memory CD4 T cells in the
presence or absence of L-tryptophan (TRP) at 24.5 µM daily and/or OX40L antibody at 10
µg/mL for 7 days. Proliferating CFSElow CD4 T cells were assessed by FACS at day 7 and the
results are shown as individual histograms. (B) Additionally, the results are shown as
means±SEM of replicates (n=2-3) for LacZ-transduced VSMC and OX40L-transduced VSMC
co-cultures, *p<0.05 (ANOVA). The data are representative of two independent experiments.
Supplemental Figure VII. OX40L is not expressed by VSMC in situ. (A) CD45, (B) CD31, and
(C) α-smooth muscle actin (α-SMA) expression in control non-aneurysmal aortas and aortic
aneurysms were determined by immunofluorescence staining. Positive labeling for these
8
leukocyte and vascular cell markers is shown in green color (left panels). OX40L expression in
the same sections is shown in red color (middle panels). Nuclei staining with DAPI is shown in
blue color. Merged images of all three colors are also shown (right panels). The vascular
compartments are labeled as A (adventitia), M (media), and I (intima). Selected areas of medial
staining from aneurysm specimen are also shown at higher magnification (insets) in addition to
lower magnification (panels). Horizontal bars in insets represent 25 µm and those in panels
represent 400 µm. Photomicrographs are representative of similar findings from six patients.
Quantification demonstrated that OX40L was not expressed in the normal aortic media and that
the great majority (>85%) of OX40L+ cells in the inflamed media of aneurysmal aortas were
CD31+ EC of angiogenic microvessels, infrequent OX40L+ cells (about 9-12%) were CD45+
leukocytes, and very few OX40L+ cells (<1%) were α-smooth muscle actin+ VSMC.
9