| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Jobst Vascular Surgery Laboratory, Section of Vascular Surgery, Department of Surgery,
Division of Magnetic Resonance Imaging, Department of Radiology, and
Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, University of Michigan Medical Center, and
§
Parke Davis Pharmaceutical Division, Ann Arbor, MI 48109; and
¶
Racanti-Miller Transplant Institute, Department of Surgery and Molecular Medicine, Mt. Sinai School of Medicine, New York, NY 10029
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The molecular pathophysiology of DVT is beginning to be delineated.
Thrombosis, like other physiologic processes, initiates a complex
inflammatory mediator-procoagulant cascade that ultimately leads to
vein wall damage (4, 5, 6, 7). The process of thrombus
resolution is similar to that of wound healing, with the nidus of
injury being the clot, followed by leukocyte influx and eventually vein
wall fibrosis. Mediators of this process include proinflammatory
cytokines such as TNF-
and IL-1ß (from the injured venous cells
and local leukocytes) that serve to further activate leukocytes,
stimulate chemotactic factor release, promote thrombosis, and increase
cell adhesion molecule (CAM) expression (4, 6, 7, 8, 9).
Furthermore, thrombin acts as a proinflammatory mediator that not only
amplifies the coagulation cascade, but also stimulates procoagulant
changes in endothelial cells with concurrently increased CAM expression
(3, 8, 10, 11, 12). The CAMs P- and E-selectin and ICAM-1 are
critical for leukocyte tethering, firm adhesion, and emigration,
respectively. Leukocyte emigration into the vein wall allows the
release of cytotoxic inflammatory intermediates that cause vein damage
(4, 6, 8, 9, 13, 14, 15). Prior work in our laboratory has
clarified that TNF- and P-selectin are central mediators of
post-thrombotic vein wall inflammation and direct the magnitude of
thrombus production (6, 13, 14, 15).
We have previously characterized an important role for IL-10 as an anti-inflammatory cytokine in the rat model of DVT. Specifically, neutralization of IL-10 resulted in increased local inflammation, while supplementation with human IL-10 (hIL-10) in a dose- and time-dependent fashion decreased not only perithrombotic inflammation but also thrombus size (5). EBV encodes a homologue to rodent and human IL-10 that is designated viral IL-10 (vIL-10). vIL-10 has similar anti-inflammatory properties as hIL-10, but may not have the lymphocyte stimulatory properties associated with hIL-10 at high concentrations (16, 17). Other investigators have found that adenovirus encoding vIL-10 (Ad-vIL-10) transfection is associated with a significant reduction in alloantigen responsiveness, significantly increased cardiac allograft survival, decreased collagen-induced arthritis inflammation, and decreased systemic TNF and IL-1ß release (18, 19, 20, 21). Adenovirus-mediated gene transfer has the advantage of high efficiency, but is limited by transient gene expression and the host immune response against the viral infection (22, 23). Theoretically, local production of IL-10 via transfection has the advantage of minimizing confounding systemic effects, and its constant local expression may further enhance the anti-inflammatory effects compared with intermittent systemic infusions. In this study we show that Ad-vIL-10 transfection in a localized venous segment confers an anti-inflammatory effect during thrombosis and that this effect is probably mediated by down-regulation of CAM expression, but not through a reduction in proinflammatory mediator or procoagulant activity.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All vectors are E1 deleted. Ad-CMV encoding ß-galactosidase (ß-gal) or vIL-10 gene were under the control of the CMV immediate early promoter. Ad-RSV-ß-gal was under the control of the RSV immediate early promoter. All vectors were obtained from the vector core of University of Michigan Medical Center and were produced as previously described (19).
Animal model, transfection, and tissue analysis
This rat model is a modification of a well-characterized method
for reproducible stasis induced venous thrombosis (3, 5, 6, 13, 24). In brief, male Sprague Dawley rats, weighing 250–300 g,
were anesthetized with an inhalation mixture of isoflurane (1–2%) and
oxygen (100%) during the procedure. On day 1, aseptic laparotomy was
performed, the inferior vena cava was isolated, and major side branches
were ligated (Fig. 1
). An 
1.2-cm
segment was dissected to allow temporary proximal and distal occlusion
with microvascular clips. After occlusion, the inferior vena cava (IVC)
was cannulated with a 30-gauge needle catheter, the blood was
aspirated, and the IVC was flushed with a dilute heparinized saline
solution. After this, 0.15-ml instillation of saline, Ad-CMV-ß-gal,
Ad-RSV-ß-gal or Ad-CMV-vIL-10 (all 5 x
108 PFU/ml final concentration) was performed for
30 min. The vector solution or saline was aspirated, the clips were
removed to re-establish blood flow, and the midline incision was
closed. Two days after the transfection, the rats were anesthetized in
the same fashion, and a repeat laparotomy was performed. The IVC was
then ligated below the renal veins for establishment of thrombosis.
This method produces a clot in >95% of animals, and gradual
re-establishment of flow usually occurs by dilation of posterior
collaterals. Two days after thrombosis induction, the animals underwent
gadolinium (Gd)-enhanced magnetic resonance imaging and were
sacrificed. The thrombosed IVC segment was harvested below the renal
veins and weighed (milligrams), and the weight was corrected to vein
length (centimeters). The vein was then halved, and the proximal
portion was placed in 10% buffered formalin for 24 h, followed by
70% ethanol for subsequent permanent section processing. The lower
segment of the vein was snap-frozen in liquid N2
and kept at -70°C until homogenized.
|
On the day of sacrifice, the rats were sedated with Telazol (20–40 mg/kg i.m.) and imaged with a magnetic resonance imaging system (Signa Horizon LX, Milwaukee, WI) using 8.1 software as previously described (24). Baseline time-of-flight imaging and T1 weighted imaging were performed followed by tail vein injection of 2 ml of Gd dimeglumine (0.2–0.4 mg/kg; Magnevist, Berlix, Wayne, NJ), and T1-weighted images were repeated. Gd is a heavy metal chelate that shortens the T1 relaxation time, prolonging signal enhancement. It extravasates into areas of tissue inflammation (capillary leak) with high sensitivity. In a blinded fashion, analysis of four or five coronal sections of the whole IVC was reviewed, the most intense Gd-enhanced segment was chosen, and the area of maximum enhancement was measured in square millimeters in treated and controls rats.
Histopathologic analysis and immunohistochemical staining
Leukocyte infiltration into the vein wall was assessed as previously described (3, 5, 6, 13). In brief, leukocyte morphometric analysis was performed in a blinded fashion on hematoxylin- and eosin-stained sections. Each rat had three sections reviewed, and the section with the most complete circumference was chosen. Then, five high power fields (x1000) radially around the clot wall interface were assessed using standard criteria for neutrophils, monocytes (Mo), and lymphocytes. Random sections were taken from the proximal and middle portions of the vein clot tissue for analysis.
Immunohistochemical analysis was performed on the paraffin-processed tissue sections (10 µm) as previously described (6). In brief, the slides were deparaffinized with xylene followed by rehydration with graded dilutions of ethanol. The sections underwent Ag retrieval using 0.1% trypsin/0.1%CaCl2 for 25 min, followed by quenching of endogenous peroxidase activity with methanol/3% H2O2 (50/50) for 15 min. On each slide after blocking nonspecific sites, the sections were incubated with either control isotype specific Ab (goat IgG) or anti-E-selectin, anti-P-selectin (both 1 µg/ml), or anti-ICAM-1 (2 µg/ml; all from Santa Cruz Biotechnology, Santa Cruz, CA) for 30 min. An anti-goat biotinylated Ab (1/1000) followed by avidin-peroxidase conjugate was added in sequence, and color development was performed with addition of diaminobenzidene substrate (ABC kit, Vector, Burlingame, CA). After color development, the slides were washed and counterstained with Mayer’s hematoxylin. Qualitative assessment of staining intensity was compared between the experimental and control groups.
5-Bromo-4-chloro-3-indolyl ß-D-galactopyranoside (X-Gal) staining was performed on the control vector sections as described to confirm transfection (18). In brief, the rat vein sections were embedded with O.C.T. (Miles Scientific, Naperville, IL), snap-frozen in liquid N2, sectioned at 10 µM, and fixed in 0.25% glutaraldehyde in PBS for 30 min. After rinsing in PBS for 30 min, the slides were incubated overnight at 37°C with the X-Gal solution (0.5 mg/ml X-Gal in 5 mM potassium ferrocyanide, 5 mM potassium ferricyanide, 2 mM MgCl2, 1 mM spermidine, 0.02 Nonidet P-40, and 0.01% sodium deoxycholate in PBS.). The sections were then fixed with 4% formaldehyde and counterstained with eosin.
Cytokine/coagulation factor/CAM analysis
The vein tissue (clot and wall) was thawed, placed in complete lysis buffer (Boehringer Mannhein, Indianapolis, IN), homogenized with a hand-held homogenizer, and sonicated for 10 s. The homogenate was centrifuged at 10,000 x g for 5 min, and the supernatant and sediment fraction were separated.
Quantification of inflammatory mediators and CAMs were all normalized
to total protein in the sample. Total protein quantitation was
performed using a modified Bradford assay according to the
manufacturer’s instructions (Pierce, Rockford, IL) with serial
dilutions of BSA (Sigma, St. Louis, MO) as standards. The analysis of
TNF- and cellular IL-10 involved pooling three rat vein tissue
segments. All other mediator analysis was performed on single rat vein
tissue samples. Tissue homogenate ELISAs for the aforementioned
mediators were quantified using a double-ligand technique as previously
described (5, 6, 13). In brief, flat-bottom 96-well
microtiter plates (Immuno-Plate I 96-F, Nunc, Naperville, IL) were
coated with 50 µl/well (1 ng/ml) of the specific rabbit
anti-cytokine Ab in coating buffer (0.6 mol/L NaCl, 0.26 mol/L
H3PO4, and 0.08 N NaOH, pH
9.6) for 16–24 h at 0°C. Nonspecific sites were then blocked with
2% BSA in PBS for 60 min at 37°C, followed by sample addition of a
50-µl aliquot in duplicate and incubated for 60 min at 37°C. After
washing, 50 µl of a biotinylated rabbit polyclonal Ab (3.5 µg/ml in
PBS pH 7.5, 0.05% Tween-20, and 2% FCS) was added and incubated for
45 min at 37°C. No Ab cross-reactivity between the cytokines was
found. Plates were washed, streptavidin-peroxidase conjugate (1/5000)
was added, and the plates were incubated for 30 min at 37°C. The
plates were again washed, the substrate TMB
(3,3',5,5'-tetramethylbenzidine, Kierkegaard & Perry, Gaithersburg, MD)
was added for color development, and the reaction was quenched using
100 µl of 1 M H2PO4.
Plates were then read at 450 nm with an automated microplate reader
(Bio-Tek Instruments, Winooski, VT). Standards were 0.5 log dilutions
of the cytokines from 1 pg/ml to 100 ng/ml with a sensitivity 
50
pg/ml.
Analysis of the rat cytokine IL-1ß and prostacyclin (estimated by
6-keto-PGF2) levels were
performed using a commercial ELISA and enzyme immunoassay,
respectively, according to the manufacturer’s instructions with a
sensitivity of 10 pg/ml (R&D Systems, Minneapolis, MN). The protein
levels of tissue factor (TF) and von Willebrand factor (vWF) were
analyzed using a human commercial ELISA with a sensitivity of 50
pg/ml or 10 milliunits/ml, respectively. (American Diagnostics,
Greenwich, CT).
P- and E-selectin and ICAM-1 levels were quantified in the cell
homogenate using a direct ELISA method. In brief, 25 µl of the vein
tissue homogenate or Ag-specific blocking peptide (for rat P- and
E-selectin and ICAM-1) as standards (serial dilutions from
0.001–10 ng/ml; Santa Cruz Biotechnology) was mixed with coating
buffer, added in duplicate, and incubated overnight at 0°C. The
plates were washed three times with PBS and blocked with 2% BSA in PBS
for 60 min at 37°C. After further washing, 50 µl of Ab to
P-selectin, E-selectin, or ICAM-1 (2 µg/ml; Santa Cruz Biotechnology)
was added and incubated for 60 min at 37°C. After washing three times
with PBS, a goat anti-rabbit peroxidase conjugate secondary Ab
(1/200 dilution; Santa Cruz Biotechnology) was added and incubated for
45 min at 37°C. The plates were again washed, TMB was added for color
development, and the reaction was quenched using 100 µl of 1 M
H2PO4. The sensitivity was
10 pg/ml, and negligible background was found when testing the
secondary Ab against the adsorbed Ags.
Analysis of procoagulant activity
An amidolytic assay was used for the determination of TF activity in the vein tissue homogenate (25). In brief, 50 µl of the vein tissue homogenate was added in duplicate to a microtiter plate followed by addition of the enzyme-substrate solution (1 U/ml Proplex T (Baxter-Allegiance Healthcare, McGaw Park, IL) and 200 µg/ml of S2222 (Helena Laboratories, Beaumont, TX) in medium 199 without phenol red (Life Technologies, Gaithersburg, MD) and with 3.24 mM CaCl2). This mixture was incubated for 30 min at 37°C in a humidified 5% CO2 incubator and then quenched with 3% HOAc. The plate was read at 410 nm. A standard of 0.01 nmol/ml of lipidated TF (American Diagnostics) was run under the same conditions, and color development was arbitrarily set at 1 µ/ml. The background of the homogenate without substrate enzyme mixture was subtracted from the sample. Negligible color development occurred in the absence of tissue homogenate.
RT-PCR
After removal of the vein tissue homogenate supernatant, the sediment fractions were pooled (n = 2–3) and placed in Trizol (Life Technologies, Gaithersburg, MD), which uses guanidinium-thiocyanate-phenol to isolate RNA. The RNA was extracted with chloroform/isoamyl alcohol and precipitated with 70% ethanol. The resulting RNA was quantified spectophometrically, and then 0.5–1 µg/ml of RNA was used to synthesize cDNA by random primer reverse transcriptase (Invitrogen, Carlsbad, CA). The cDNA was amplified using two specific primers for the vIL-10 gene as previously described (18, 26): 5'-ATGGAGGAGCGAAGGTTAGTGGTC-3' (upstream) and 5'-ACTCTTGTTCTCACACGGCAG-3' (downstream). Forty PCR cycles yielded a 387-bp fragment using an automated thermocycler (RJ Research, Swedesboro, NJ) with cycles of 96°C for 15 s, 60°C for 30 s, and 72°C for 2 min. The cDNA from a vIL-10 plasmid-transfected COS cell line using the same RT-PCR reaction was used as a positive control (26). ß-Actin was amplified from the same cDNA in a separate reaction using published sequences and the same PCR cycle settings. The product was separated on a 2% agarose gel and visualized with 1% ethidium bromide staining with a UV source.
Statistical analysis and animal care
Statistical evaluation included data presentation as the
mean ± SEM. One-way ANOVA with Tukey’s highest significant
difference test for pairwise comparison or unpaired Student’s
t test was used as appropriate. Statistical significance was
assigned for p 
0.05. All experiments were repeated
at least twice. All animals in this study were housed and cared for at
the University of Michigan Unit for Laboratory Animal Medicine under
the direction of a veterinarian in accordance with the Guide for the
Care and Use of Laboratory Animals (National Institutes of Health
Publication 86-23, revised 1985).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The rat IVC was isolated, and transfection with the vector
encoding ß-gal or vIL-10 before thrombus induction was performed.
X-Gal staining on Ad-LacZ control IVCs (n = 4)
representing two separate transfection experiments demonstrated that
about 20–30% of the inner vein wall circumference stained positive
for ß-gal expression (Fig. 2,
A and B). RT-PCR analysis (pooled groups of three
separate rat vein prepared segments; n = 3 separate
transfection experiments) revealed that only the vIL-10-transfected
group produced a 387-bp fragment consistent with vIL-10 (Fig. 2C). Neither the saline nor ß-gal controls had any bands
when RT-PCR was performed using the specific vIL-10 primers. However,
all groups had similar intensity bands of the housekeeping gene
ß-actin, suggesting adequate starting cDNA.
|
Assessment of inflammation was quantified by leukocyte vein wall
emigration and Gd-enhanced MRV. Total subendothelial and medial
leukocytes were all decreased with vIL-10 transfection compared with
those in control saline and ß-gal rats (n = 6–12;
Fig. 3). Specifically, total leukocytes
in the vein wall were significantly reduced by 36% (vIL-10, 33 ±
3; ß-gal, 47 ± 4; saline, 56 ± 8), while neutrophils were
reduced by 45% (vIL-10, 16 ± 3; ß-gal, 27 ± 4; saline,
31 ± 6) compared with controls (p
0.05, by ANOVA, vIL-10 vs saline and ß-gal; p <
0.05, by t test, vIL-10 vs. ß-gal). Similarly, Mo
extravasation was reduced by 27% (p < 0.05),
and lymphocytes in the vein wall were rare in all groups.
|
). Coronal venous section analysis of
the ß-gal and saline control rats revealed 31 ± 4- and 31
± 5-mm2 average areas of perithrombotic Gd
enhancement, respectively. In contrast, the rats transfected with
vIL-10 had a decreased average coronal vein section area of
perithrombotic enhancement of 17 ± 2 mm2
(p = 0.056, by ANOVA, vIL-10 compared with
saline and promoter controls; p < 0.05 by t
test, vIL-10 compared with promoter control). Whereas the Gd
enhancement was fairly diffuse in the control groups, the zone of
enhancement was tighter and seemed to be mainly associated with the
clot in the vIL-10 group. These data reflect a decreased amount of
capillary leak and therefore decreased inflammation in the
vIL-10-transfected group. No significant enhancement was observed
before Gd injection by review of baseline scans (data not shown), and
from previous work, nonclotted IVC has an average Gd enhancement of 1.8
mm2 (24).
|
ELISA analysis of the vein tissue homogenates showed a trend
toward a decreased TNF- level in the vIL-10 group compared with
those in ß-gal and saline controls, but this did not reach
statistical significance (n = 2–4 groups of three
pooled vein segments; Fig. 5A). In contrast, IL-1ß was
not significantly different between the groups (n =
5–6; p > 0.05, by ANOVA). To test that endogenous rat
IL-10 was not accounting for the differential effects observed, an
ELISA for murine IL-10 that does not cross-react with vIL-10 was used.
Indeed, no significant differences were found in local IL-10 levels,
although a trend toward lower concentrations was found in the
transfected animals (n = 2–4 groups of three pooled
animals; saline, 918 ± 43 pg/mg protein; ß-gal, 569 ± 185
pg/mg protein; vIL-10, 604 ± 90 pg/mg protein).
|
, and is an antithrombotic mediator and an indirect
marker of inflammation. Interestingly, PGI2
levels were significantly higher in the vIL-10-transfected group
compared with those in the ß-gal control (47 ± 10 vs 21.9
± 3 ng/mg protein; p 0.05, by t test;
Fig. 5
B) and was not significantly
(p = 0.3) different from that in the saline
control (n = 5–6). Viral IL-10 transfection is associated with decreased CAM levels
We focused on endothelial-luminal-clot border staining, as
leukocytes probably migrate inward from the retracted clot-vein wall
interface. By 2 days postligation, about one-third of the clot is
retracted from the vein wall, and some blood flow around and in the
clot is present (from dilated collaterals and channels).
Immunohistochemistry was performed on permanent tissue sections to
qualitatively assess CAM intensity between the groups
(n = 4). These data suggested qualitatively less
intense luminal and medial staining of P- and E-selectin (Fig. 6). Low intensity staining of ICAM-1 was
noted in all groups, with no obvious difference seen between the groups
(data not shown). Importantly, background staining was negligible and
carefully controlled with each section (data not shown). Some positive
medial staining of P- and E-selectin in leukocytes was noted in all
groups, suggesting possible internalization of Ag by the emigrating
cells. Vein tissue section hematoxylin and eosin staining not only
yielded the already discussed quantitative leukocyte counts, but also
spatially revealed leukocytes in various stages of emigration, again
with greater numbers in the saline and ß-gal control groups compared
with the vIL-10-transfected rats.
|
). Specifically, E-selectin was most
markedly reduced (nearly 14-fold), while P-selectin was reduced 4-fold
compared with levels in the control groups (p
0.05). ICAM-1 was also significantly reduced (3-fold) compared with
the controls (p 0.05). However, this
analysis does not take into account which cell types specifically have
reduced CAM expression. Nonclotted vein tissue homogenates had
negligible levels of CAMs (data not shown), reflecting low basal
expression as well as the fact that no clot (or inflammatory nidus) was
present.
|
Clot weight normalized to length was not reduced by vIL-10
expression (saline, 79 ± 8 mg/cm; ß-gal, 65 ± 7 mg/cm;
vIL-10, 65 ± 7 mg/cm; n = 10–15). No difference
in vein-clot gross morphology was found between the groups. In
accordance with the anatomical findings, functional TF activity, TF
protein levels, and vWF levels were not significantly different in the
vIL-10 group compared with the controls (n = 3–5; Fig. 8). Nonclotted vein wall analysis showed
2- to 3-fold less TF activity and Ag levels (data not shown). Thus, no
direct difference in functional or biochemical measures of tissue
coagulant activity was found with vIL-10 expression or adenovirus
transfection.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In this animal model of venous thrombosis, we found that local expression of vIL-10 decreased inflammation. This finding correlates well with the results of our previous study in which hIL-10 given at time of thrombus induction significantly reduced inflammation (5). The absolute reduction in leukocytes (and the differential) was similar whether IL-10 was intermittently administered or constantly expressed, as presented here, suggesting a limit for the maximal anti-inflammatory effect. That no significant difference in rat cellular IL-10 levels was found between the groups suggests that the observed anti-inflammatory effect was specifically mediated by vIL-10 and not by endogenous IL-10 up-regulation. Indeed, the transfected control and vIL-10 rats had lower native IL-10 levels than saline controls. Although the local vIL-10 tissue levels were not directly determined, an unquestionable effect was seen by two parameters of inflammation, vein wall leukocyte emigration and Gd-MRV. It is important to note that the aforementioned measures of inflammation correlate well with other measures of inflammation, including Evan’s blue dye extravasation (capillary leak) and myeloperoxidase activity (marker of polymorphonuclear cells (PMN)) (4, 5, 6, 13). Also congruent with our prior study was the finding that vIL-10 reduces PMN extravasation to the greatest extent of all the leukocytes. Neutrophils are the most prevalent early leukocyte after thrombosis and are probably a primary mediator of vein wall damage (5, 6, 7).
Other investigators have found similar anti-inflammatory efficacy with vIL-10 transfection in various experimental settings. For example, the collagen-induced arthritis rat model is similar to our model, in that a chronic inflammatory stimulus is present. With Ad-vIL-10 transfection, significantly decreased arthritic inflammation compared with that in controls was observed (20). Other animal models in which an inflammatory nidus is not present also support the thesis that local vIL-10 expression is anti-inflammatory. Specifically, in a murine cardiac allograft transplant model, retrovirally mediated vIL-10 transfection improved allograft survival by 3-fold (18). In a similar model using adenovirus-mediated vIL-10 transfer, vIL-10 expression accounted for a 2.5-fold increase in the duration of allograft survival (19). The authors also suggest that incorporating the vIL-10 gene may decrease the host immune response to the adenovirus itself, a significant problem with this method of gene transfer (22, 23, 27, 28, 29). The anti-inflammatory effect of vIL-10 in these studies was diminished T cell responsiveness and cytotoxicity. We did not specifically look at these T cell parameters, but it is likely that this effect plays only a small role in the diminished inflammation that was observed because lymphocytic infiltration is quantitatively a small component of the total vein wall leukocytes. We also did not find any evidence of increased inflammation in the Ad-ß-gal control group, perhaps because a significant inflammatory stimulus was already present and "hid" any additional inflammatory effect of the viral transfection. Also, at 4 days after transfection an adoptive immune response is not expected.
Viral IL-10 decreases CAM up-regulation
Cell adhesion molecule up-regulation occurs in the setting of multiple biologic events, such as wound healing, atherosclerosis, and cancer. Indeed, the regulation of CAMs is very complex and controlled in part by the balance of pro- and anti-inflammatory cytokines (4, 8, 9). However, while the specific stimuli for increased CAM expression are fairly well defined, inhibitors of expression are less well known. In the setting of thrombosis, increased CAM expression is seen at the onset of thrombus induction and facilitates stabilization of the fibrin-platelet leukocyte complex as well as allows the proper recruitment of leukocytes for clearance and resolution of the inflammatory process (8, 9, 11, 12). Most experimental data of CAM kinetics have been made in flow situations, whereas it is less clear what occurs in a blood stasis model. Furthermore, the kinetics and response of the CAMs are different when comparing in vitro to in vivo situations. For example, P-selectin is rapidly mobilized after a stimulus such as thrombin, whereas E-selectin expression increases several hours later (4, 8). However, in vivo, persistent up-regulation of both P and E selectin may occur, depending on the inflammatory stimulus (30, 31). In the setting of venous thrombosis, an inflammatory stimulus (clot) is present that modulates CAM up-regulation, with P-selectin present in greatest magnitude.
Although we have only shown an association, our data strongly argue
that vIL-10 mediates its anti-inflammatory action through
down-regulation of CAM expression. In support of this contention, other
investigators have found that IL-10 mediates some of its
anti-inflammatory activity through decreased CAM expression.
Specifically, in an acute lung injury model, IL-10, but not IL-4,
significantly decreased lung injury (hemorrhage and myeloperoxidase
activity), and this was associated with decreased ICAM-1 levels,
probably due to concurrently decreased TNF- levels
(32). An IL-10 knockout mouse model further enforces the
close relationship between IL-10 and selectin regulation. After a low
grade septic insult, the IL-10 knockout animals had markedly increased
leukocyte adhesion in mesenteric venules and increased capillary leak
and myeloperoxidase activity compared with wild-type animals.
Importantly, when anti-P- and anti-E-selectin Abs were given,
reduction of the leukocyte adherence was observed, suggesting
unbalanced CAM up-regulation that was not seen in the wild-type mice
(33). Even more convincingly, administration of either Ab
to P-selectin or a specific soluble P- and E-selectin antagonist,
P-selectin glycoprotein ligand-1, significantly reduced inflammation
and thrombus, as measured biochemically, anatomically, and by Gd-MRV in
a non-human primate model of venous thrombosis (14, 15).
Both of these models involved experimental analysis beyond 2 days,
suggesting that early inhibition of the selectin-mediated events
may confer later anti-inflammatory effect and preserve vein wall
integrity. Other models support the interrelation between IL-10 and
CAMs. IL-10 similarly blunted inflammation secondary to myocardial
ischemia-reperfusion through an ICAM-1 mechanism (34),
while in a mouse septic shock model, IL-10 reduced liver injury and
mortality through an ICAM-1 and VCAM mechanism (35).
In vitro cell culture systems have yielded some conflicting insights into the modulatory action of IL-10 on CAM expression. In an LPS-stimulated glial cell line, IL-10 significantly reduced the expression of ICAM-1 but not its message level (36). Similarly, an IL-1-stimulated human umbilical vascular endothelial cell line had ICAM-1 and VCAM expression significantly reduced by IL-10 in a dose-dependent manner and correlated with diminished Mo and lymphocyte adhesion (37). In contrast, IL-10 administration in a human microvascular endothelial cell line actually stimulated E-selectin expression (38). These data suggest possibly a direct effect of IL-10 on endothelial CAM expression. However, experiments with human umbilical vascular endothelial cells in our laboratory have failed to show any direct modulating effect of IL-10 (over a wide concentration range) on E-selectin, ICAM-1, or procoagulant activity (unpublished observations). These results underscore the important functional heterogeneity of endothelial cells as well as the local environmental influences (4, 7, 8). As such, we believe that vIL-10 probably works indirectly on CAM expression.
The observation that P-selectin localizes rapidly and specifically to the endothelial border after thrombin stimulation underlies the importance of its role in the initial clot formation and subsequent inflammation (12, 39). Our immunohistochemical data suggest decreased luminal expression of the selectins in the vIL-10-transfected animals, and we hypothesize that luminal expression of the selectins is most important for leukocyte emigration. IL-10 may decrease the CAM expression on other cell types besides endothelial cells, and our results with the homogenate ELISA were not able to differentiate specific cell type levels. For example, IL-10 exerts an inhibitory effect on Mo ICAM-1 expression (40) that, again, may confer an anti-inflammatory effect by counterligand inhibition.
Viral IL-10 and its relationship to other inflammatory mediators
IL-10 is a pleiotropic cytokine derived primarily from Mo and T
lymphocytes that has many other anti-inflammatory actions besides
CAM modulation and may also account for the decreased inflammation
observed in these studies. Importantly, IL-10 inhibits Mo Ag-presenting
capacity and IL-1ß and TNF- production, as well as increases IL-1R
antagonist levels and lymphocyte IL-4 production (16, 17, 41). Furthermore, the neutrophils in the vein wall may have a
diminished release of reactive oxygen intermediates, while Mo nitric
oxide and chemotactic chemokines are reduced by IL-10 (42, 43). TNF- is an important proximal stimulus for CAM
up-regulation (4, 8, 9), and although we only found a
trend in decreased TNF- in the vIL-10-transfected animals, this may
be physiologically important. Two hypotheses may account for the
nonsignificant differences found in TNF-. First, the clot
environment is more static than that of other inflammation models;
thus, the TNF- Ag may not be cleared as readily (falsely elevated
local levels). Secondly, the measured TNF- levels may not correlate
with functional activity. Alternatively, other inflammatory mediators,
such as C5a, IL-6, and IFN-
, may also be down-regulated by IL-10 and
indirectly account for the decreased CAM levels (16, 17).
Interestingly, PGI2 was significantly less in ß-gal compared with vIL-10 animals. PGI2 is a vasodilatory and antithrombotic PG that is adaptively up-regulated in chronic thrombosis (44). IL-1ß levels correlated with PGI2 levels (although not significantly different between the groups), and it is well documented that IL-1ß is a proximal stimulus for PGI2 production (45, 46). Thus, it seems that vIL-10 expression preserves and may stimulate the antithrombotic PGI2 response, and that adenovirus transfection inhibits this compensatory response to chronic thrombosis compared with that in saline controls. Furthermore, our in vivo data are supported by in vitro experiments showing that IL-1ß-stimulated PGI2 production is dissociated from vWF production (47). Additionally, PGI2 has anti-leukocyte adhesive properties (48) that may have contributed to the observed decrease in vein wall leukocytes in addition to the decreased CAM expression.
Viral IL-10 does not affect tissue coagulation mechanisms
The interrelationship between CAM up-regulation and thrombosis is well defined (7, 8, 9, 11, 12, 13, 14, 15). P-selectin may be particularly important, as it directly links platelets, leukocytes, and endothelial cells to propagate thrombus. Our prior experiments support the interrelation between inflammation and thrombus production, as exogenous IL-10 decreased clot weight compared with control values (5). In the current study, however, expression of vIL-10 did not inhibit thrombus production as assessed by clot weight, nor did it affect any of the clotting parameters measured. This might relate to a difference in receptor binding constants between vIL-10 and endogenous cellular IL-10 despite the high sequence homology between these molecules (41). IL-10 decreases Mo TF activity (49), and thus, systemic administration may more fully deactivate this than when vIL-10 is localized at the thrombus site. This hypothesis is also supported by the fact that only IL-10 given at the time of thrombus induction was an effective antithrombotic agent (5). It is more likely that adenovirus transfection itself stimulated the procoagulant state that was not overcome by vIL-10. Adenovirus transfection (with or without ß-gal expression) in normal rabbit arteries is associated with a significant inflammatory response, increased CAM expression, and the development of neointimal hyperplasia compared with that in nontransfected cells (22). The host inflammatory response to adenovirus infection may occur even in the absence of gene expression (27) and does not appear to be specifically related to ß-gal expression (22). This inflammatory response may be reduced by adventitial rather than luminal delivery of the vector (29). Certain viral vectors may also promote procoagulant changes in transfected endothelial cells (28). We did not find that either procoagulant activity or TF Ag was significantly elevated in the ß-gal control transfected animals, but this may be because the maximum thrombus and local inflammatory milieu stimulus was already present.
Summary
Local expression of vIL-10 effectively reduces early vein wall inflammation, probably by indirect down-regulation of P- and E-selectin and ICAM-1 expression. The important question remains, however, as to whether this reduces late vein wall damage or affects the development of chronic venous insufficiency. Additional experiments to better delineate the differences between vIL10 and rIL10 are planned by systemic administration, which might abrogate differences in gene expression variability. Using specific CAM knockout mice would also clarify the actions of vIL10 in this system. Although our methodology of segmental transfection of an anti-inflammatory mediator was used here as a tool to delineate its effect and mechanisms, clinical applicability is foreseeable. However, before clinical use of such an anti-inflammatory cytokine, further physiologic experimental measures, such as distal femoral venous pressure, ultrasound analysis of clot resolution and vein wall thickness, and vein wall analysis for fibrosis, should be undertaken. Furthermore, longer duration transfection studies to determine the durability of the effect of anti-inflammatory cytokines would be important before any clinical use. Given the unremitting nature of chronic venous insufficiency (CVI) and the increased likelihood of rethrombosis in an anatomically diseased segment, local gene-specific anti-inflammatory or anti-coagulant mediators are appealing. The complications from long term anticoagulation are significant (1), and selectively preventing vein wall rethrombosis and/or inflammatory damage without systemic anticoagulation is important and worth further investigation.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 A portion of this study was presented at the 85th Annual Meeting of the American College of Surgeons-Surgical Forum, October 14, 1999, San Francisco, CA.
3 Address correspondence and reprint requests to Dr. Peter Henke, Section of Vascular Surgery, Department of Surgery, 2210 Taubman Health Care Center, 1500 East Medical Center Drive, Ann Arbor, MI 48109-0329. E-mail address:
4 Abbreviations used in this paper: DVT, deep vein thrombosis; CAM, cell adhesion molecule; hIL, human IL; vIL-10, viral IL-10; Ad-vIL-10, adenovirus encoding vIL-10; ß-gal, ß-galactosidase; RSV, Rous sarcoma virus; IVC, inferior vena cava; MRV, magnetic resonance venography; Mo, monocytes; TF, tissue factor; vWF, von Willebrand factor; PMN, polymorphonuclear cells; X-Gal, 5-bromo-4-chloro-3-indolyl ß-D-galactopyranoside; TMB, 3,3',5,5'-tetramethylbenzidine.
Received for publication August 2, 1999. Accepted for publication December 8, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and IL-1ß, not IL-6, in the newborn endotoxemic mouse. J. Pediatr. Surg. 31:411.[Medline]
This article has been cited by other articles:
|
|
 |
|
  E. Cochery-Nouvellon, P. Nguyen, R. Attaoua, P. Cornillet-Lefebvre, E. Mercier, F. Vitry, and J.-C. Gris Interleukin 10 Gene Promoter Polymorphisms in Women with Pregnancy Loss: Preferential Association with Embryonic Wastage Biol Reprod, June 1, 2009; 80(6): 1115 - 1120. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I.-S. Shin, J.-M. Kim, K. L. Kim, S. Y. Jang, E.-S. Jeon, S. H. Choi, D.-K. Kim, W. Suh, and Y.-W. Kim Early growth response factor-1 is associated with intraluminal thrombus formation in human abdominal aortic aneurysm. J. Am. Coll. Cardiol., March 3, 2009; 53(9): 792 - 799. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. K. Henke, C. G. Pearce, D. M. Moaveni, A. J. Moore, E. M. Lynch, C. Longo, M. Varma, N. A. Dewyer, K. B. Deatrick, G. R. Upchurch Jr, et al. Targeted Deletion of CCR2 Impairs Deep Vein Thombosis Resolution in a Mouse Model. J. Immunol., September 1, 2006; 177(5): 3388 - 3397. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. W. Raines and N. Ferri Thematic Review Series: The Immune System and Atherogenesis. Cytokines affecting endothelial and smooth muscle cells in vascular disease J. Lipid Res., June 1, 2005; 46(6): 1081 - 1092. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Lucas and G. McFadden Secreted Immunomodulatory Viral Proteins as Novel Biotherapeutics J. Immunol., October 15, 2004; 173(8): 4765 - 4774. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. K. Henke, A. Varga, S. De, C. B. Deatrick, J. Eliason, D. A. Arenberg, P. Sukheepod, P. Thanaporn, S. L. Kunkel, G. R. Upchurch Jr, et al. Deep Vein Thrombosis Resolution Is Modulated by Monocyte CXCR2-Mediated Activity in a Mouse Model Arterioscler. Thromb. Vasc. Biol., June 1, 2004; 24(6): 1130 - 1137. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. H. Von der Thusen, J. Kuiper, T. J. C. Van Berkel, and E. A. L. Biessen Interleukins in Atherosclerosis: Molecular Pathways and Therapeutic Potential Pharmacol. Rev., March 1, 2003; 55(1): 133 - 166. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Daddi, T. Suda, F. D'Ovidio, S. A. Kanaan, T. Tagawa, K. Grapperhaus, B. D. Kozower, J. H. Ritter, N. S Yew, T. Mohanakumar, et al. Recipient intramuscular cotransfection of naked plasmid transforming growth factor {beta}1 and interleukin 10 ameliorates lung graft ischemia-reperfusion injury J. Thorac. Cardiovasc. Surg., August 1, 2002; 124(2): 259 - 269. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Yang, M. Chen, R. Wu, L. B. Fialkow, J. S. Bromberg, M. McDuffie, A. Naji, and J. L. Nadler Suppression of Autoimmune Diabetes by Viral IL-10 Gene Transfer J. Immunol., June 15, 2002; 168(12): 6479 - 6485. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. A. Gunnett, D. D. Heistad, and F. M. Faraci Interleukin-10 Protects Nitric Oxide-Dependent Relaxation During Diabetes: Role of Superoxide Diabetes, June 1, 2002; 51(6): 1931 - 1937. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. M. Carlos Leukocyte recruitment at sites of tumor: dissonant orchestration J. Leukoc. Biol., August 1, 2001; 70(2): 171 - 184. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Shen, N. Taylor, L. Duncan, I. Kovesdi, J. T Bruder, J. V Forrester, and A. D Dick Ex vivo adenovirus mediated gene transfection of human conjunctival epithelium Br. J. Ophthalmol., July 1, 2001; 85(7): 861 - 867. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Sewnath, D. P. Olszyna, R. Birjmohun, F. J. W. ten Kate, D. J. Gouma, and T. van der Poll IL-10-Deficient Mice Demonstrate Multiple Organ Failure and Increased Mortality During Escherichia coli Peritonitis Despite an Accelerated Bacterial Clearance J. Immunol., May 15, 2001; 166(10): 6323 - 6331. [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]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
