| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Division of Cardiothoracic Surgery, Department of Surgery, and 2Division of Cardiology, Department of Medicine, University of California Los Angeles Medical Center, and David Geffen School of Medicine in University of California Los Angeles, Los Angeles, California 90095
Submitted 13 November 2002 ; accepted in final form 2 May 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-converting enzyme) and activated caspase-9 were
increased 7- and 2.3-fold, respectively, in IL-10 gene-treated allografts at
postoperative day 4–5. These changes in the Fas-Fas ligand
pathway and Bcl-2 mitochondrial apoptosis regulation are enhanced by complete
suppression of antiapoptotic FADD-like IL-1-converting enzyme inhibitory
protein (FLIP) (from 30.5 to 0.0%, P < 0.01) and Bcl-xL (from 22.5
to 0.1%, P = 0.03) expression among these cells from the earliest
days after gene transfer. Although changes in proteins of Fas- and
Bcl-2-mediated apoptosis signaling occur, only the levels of Fas and FLIP
correlate to the rate of apoptosis of graft-infiltrating CD3 lymphocytes and
histological rejection scores. These results indicate that dichotomous
apoptosis-regulatory pathways are affected by IL-10 gene therapy, but
Fas-mediated mechanisms of activation-induced cell death more substantially
contribute to the greater cell death of graft-infiltrating T cells after ex
vivo IL-10 gene transfer.
activation-induced cell death; T lymphocytes; cardiac allograft rejection; gene therapy; interleukin-10
A necessary adjunct to CD3-TCR complex stimulation that prevents premature
AICD of activated T cells is costimulation by professional antigen-presenting
cells (APC) (40). Studies
demonstrate that stimulation at CD28 prevents apoptosis of previously
activated T cell populations, and this protection is mediated through
increased expression of Bcl-xL and FADD-like IL-1-converting enzyme
(FLICE) inhibitory protein (FLIP)
(6,
26). FLIP acts to block
activation of procaspase-8 by CD95 and its associated intracellular adapter
molecules (23), whereas Bcl-xL
coordinately regulates apoptosis at the mitochondrial membrane surface with
other members of the Bcl-2 protein family
(1). Thus restimulation of
activated T cells through the CD3-TCR complex in the absence of costimulation
could trigger AICD, whereas costimulation might allow these cells to continue
an ongoing effector function.
In addition to costimulation, AICD is also modulated by cytokines. As previously discussed, IL-2 is critical for sensitizing T cells to AICD, and its congenital absence in "knockout" mice causes accumulation of peripheral lymphocytes and autoimmunity (27). Additionally, the Th1 cytokine IL-12 reduces AICD of activated T cells in culture (13). IL-10 is a pleiotrophic cytokine produced by Th2 T cells that acts to inhibit the Th1 response through reduction of Th1 cytokine secretion and blockade of APC function with downregulation of myosin heavy chain class II and B7 expression (18, 31). In cell culture, exogenous IL-10 has minimal effect on AICD of preactivated T cells (13). However, inhibition of endogenous IL-10 through addition of monoclonal IL-10 antibody to these cells greatly diminishes AICD on TCR restimulation (5, 13). Although the effect of IL-10 on AICD could potentially be mediated through downregulation of costimulatory molecules such as B7, recent studies also show that IL-10 has a direct inhibitory influence on intracellular signaling by CD28 through reduction of tyrosine phosphorylation at its cytoplasmic tail (2).
To further investigate the effect of IL-10 on T cell apoptosis, we employed an in vivo model using heterotopic cardiac transplantation with gene transfer to overproduce IL-10 in an environment with continuous high-level immunologic T cell stimulation. With this model, we quantified the extent of apoptosis of infiltrating T cells and the concomitant expression of apoptosis regulatory molecules in situ to better understand the cellular mechanisms by which IL-10 regulates T cell apoptosis.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
strain of Escherichia coli (GIBCO-BRL,
Carlsbad, CA), extracted by alkaline lysis, and purified according to the
Qiagen Endofree plasmid purification protocol. The purity and identity of the
DNA samples were confirmed by absorbance measurements and by agrose gel
electrophoresis. Endotoxin levels were reduced to <0.1 endotoxin unit/mg
plasmid DNA. The cationic liposome was GAP:DLRIE (GIBCO-BRL), a new generation
in the 2,3-dioxy-propaniminium class of cationic lipid basic skeleton that
also includes
(+)-N-(2-hydroxyethyl)-N,N-dimethyl-2,3-bis(tetradecyloxy)-1-propaniminium
bromide (DLRIE),
N-[1-(2,3-dioleyloxy)propyl]-N,N,N-trimethyl ammonium,
1,2-bis(oleoyloxy)-3-(trimethyl-ammonio)propane, and
2,3-dioleyloxy-N-[2-(sperminecarboxamido)ethyl]-N,N-dimethyl-1-propanammonium.
This class of cationic lipids has two hydrophobic chains appended to a
quaternary ammonium moiety via a polar dioxy-propyl group to create a central
glycerol-like structure. Recombinant IL-10 cDNA plasmid (50 µg) was
complexed with this liposome (75 µg), and the mixture was vortexed for 20
min just before use. The optimized mean ex vivo liposome-mediated IL-10 gene
transfection rate in hearts was 15 ± 3%, which is 35 times higher than
the rate with naked plasmids but >4 times lower than the rate with
adenoviral vectors. With this model, IL-10 gene expression peaks at
postoperative day (POD) 8 to reach a steady state,
and IL-10 protein expression has equivalent expression throughout the four
chambers of the heart (22,
32,
36).
|
Functional heterotopic heart transplant model with ex vivo gene transfer. These studies employed a functional cervical heterotopic heart transplant model that has recently been described by our laboratory in rabbits (32, 36). This model more closely resembles normal physiology than the more widely used nonfunctional vascularized heterotopic transplants, because 1) the left heart is filled with only oxygenated blood and 2) both ventricles receive sufficient preload.
All animals received humane and ethical care according to the guidelines of the National Institutes of Health and the Association for Assessment and Accreditation of Laboratory Animal Care. Donors and recipients were New Zealand White rabbits purchased from geographically unrelated vendors and weighed 3.5 kg (Charles River Laboratory) and 4.0 kg (Irish Farms, Norco, CA), respectively. No histocompatability matching was performed, but rabbits from each vendor were third-generation sibling-inbred male animals. During surgery, the animals were sedated with intravenous ketamine (10 mg/kg) and acepromazine (1 mg/kg) before endotracheal intubation and maintenance of anesthesia with 1.5–2.0% inhaled isoflurane. After administration of heparin (200 IU/kg) to the donor, a cannula was inserted into the ascending aorta through the right carotid artery for cardioplegia and gene infusion. Systemic venous return was then occluded by ligation of the superior and inferior venae cavae, and the pulmonary artery was transected. After the aorta was clamped distal to the left carotid artery, antegrade coronary perfusion with University of Wisconsin solution (4°C, 100 ml) was begun simultaneously with application of topical ice slush. The left atrium was then opened for ventricular decompression. The pulmonary veins were ligated, and the heart was then excised and placed in cold (4°C) University of Wisconsin solution. Next, liposome IL-10 gene complexes were administered through ex vivo intracoronary infusion at 20 ml/h. Control animals received empty liposome vehicle instead of gene. Donor hearts were then heterotopically transplanted into rabbits in the right cervical position. The aorta and left atrium of the donor heart were anastomosed to the recipient's right carotid artery. The pulmonary artery and right atrium were anastomosed to the right jugular vein. Topical cooling of the donor heart was maintained throughout the procedure. All allografts resumed spontaneous, vigorous contraction within minutes of reperfusion. Graft function was thereafter assessed twice daily by palpation and auscultation of the right neck. Graft failure was defined as loss of palpable beating and heart sounds.
RT-PCR for hIL-10 transgene expression. Functioning allografts
were retrieved from recipient rabbits after euthanasia on POD 1–2,
4–5, 7–8, and 11–12 in the gene therapy group
and POD 1–2, 4–5, and 7–8 in the control
group. Total RNA was isolated from this tissue by the lithium chlorideurea
method. cDNA was synthesized from total RNA by random primer RT using 200
units of SuperScript II reverse transcriptase (GIBCO-BRL). The RNA-primer
mixture was incubated at 42°C for 60 min and then at 70°C for 15 min.
cDNA was amplified with 1 unit of Taq DNA polymerase (GIBCO-BRL) and
0.2 µM primer to a final volume of 100 µl. The primers
[5'-ATGGAGCGAAGGT-TAGTGGTCA-3' (sense) and
5'-CTCGCTTTAATTGTCATGTATGCT-3'(antisense)] amplified a 461-bp
region of the IL-10 cDNA. The -actin housekeeping gene was coamplified
for each experiment (Fig.
1B). Thirty-five cycles of amplification were performed
in a thermocycler (GeneAmp PCR System 9600, Perkin-Elmer, Wellesley, MA) and
consisted of denaturing at 94°C for 30 s, primer annealing at 55°C for
45 s, and primer extension at 72°C for 30 s. The reaction mixtures were
preheated at 94°C for 2 min before amplification. An extra 10 min at
72°C was added for extension at the conclusion of the PCR cycles. RT-PCR
products were analyzed by electrophoresis on 1.2% agarose gels and visualized
with ethidium bromide staining (10 mg/ml). Band densities were evaluated by
densitometric scanning (NIH Image Software, NIH, Bethesda, MD).
ELISA for IL-10 protein expression. Fresh ventricular tissue was
homogenized in 50 mM tromethamine chloride, 150 mM sodium chloride, 5%
-mercaptoethanol, 2 mM EDTA, and 1 mM phenylmethylsulfonyl fluoride,
with 10 µg/ml N-tosyl-L-phenylalanine chloromethyl
ketone, 10 µg/ml N--tosyl-L-lysine chloromethyl
ketone, 5 µg/ml E64, 5 µg/ml antipain, and 25 µg/ml leupeptin. The
amount of hIL-10 was quantified in these extracts using the standard
Pharmingen protocol for sandwich ELISA. 96-Well flat-bottom plates were coated
with 50 µl of monoclonal mouse anti-hIL-10 antibody (BD Pharmingen, San
Diego, CA) overnight at 4°C, washed twice with phosphate-buffered saline
(PBS)-Tween solution, and blocked with PBS containing 10% fetal calf serum for
1 h. After the plates were washed with PBS-Tween, IL-10 standards, myocardial
extract samples, and positive and negative controls were added for 2 h at room
temperature. Secondary biotinylated monoclonal mouse anti-hIL-10 antibody (100
µl; BD Pharmingen) was then added at room temperature for 45 min.
H2O2 (0.1%) in
2,2'-azinobis-(3-ethylbenzthiozolinesulfonic acid) substrate was used
for detection, and quantification was performed on an automated ELISA plate
reader (Dynatech Laboratories, Middlesex, UK) at 450 nm. Results for each
sample were compared with the standard curve and expressed in nanograms per
gram of allograft tissue.
Hematoxylin-eosin staining and histological evaluation. Left ventricular tissues were embedded in ornithine carbamoyltransferase and rapidly frozen on dry ice. Serial frozen sections were cut (5 µm thick). One section was stained with hematoxylin and eosin. A third-party observer blinded to the treatment group origins of the tissue then assigned a rejection score on the basis of the standardized nomenclature used in the diagnosis of cardiac allograft rejection established by the International Society for Heart and Lung Transplantation.
In situ identification of T cells. To study the extent of T cell infiltration, additional sections were stained for the cell surface markers CD3, CD4, and CD8. These sections were fixed in 100% acetone at 4°C for 20 min and rehydrated in PBS. Endogenous peroxidase activity was then quenched with 3% H2O2, and nonspecific binding was blocked with 1% bovine serum albumin (BSA) for 20 min. Sections were then incubated with biotinylated mouse monoclonal antibodies raised against rabbit CD3, CD4, or CD8 (Spring Valley Laboratories, Woodbine, MD) for 1 h at room temperature. Antibody-biotin conjugate was detected with 3-amino-9-ethylcarbamazole (AutoProbe III Immunhistochemistry Kit, Biomeda, Foster City, CA) according to the manufacturer's instructions. Slides were counterstained with hematoxylin. Positive cells stained orange-brown with this method.
In situ identification of nuclear DNA fragmentation. In situ terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick end labeling (TUNEL) of tissue sections was used to detect apoptosis. Cryopreserved serial sections adjacent to those stained for CD3, CD4, or CD8 were used for these assays. After fixation with 4% formaldehyde in PBS, these sections (5 µm thick) were permeabilized with proteinase K (20 µg/ml in 10 mM Tris · HCl, pH 8.0) for 10 min at room temperature. After sections were rinsed in Tris-buffered saline (20 mM Tris, pH 7.6, 140 mM NaCl), endogenous peroxidases were inactivated with 3% H2O2 in methanol. Sections were then incubated at room temperature with TdT equilibration buffer (Oncogene Research, Cambridge, MA) for 30 min and then for 90 min at 37°C with TdT enzyme (1 U) and biotinylated nucleotides. The reaction was terminated by incubation with 0.5 M EDTA, pH 8. The sections were then blocked with 4% BSA in PBS, incubated for 30 min at room temperature with peroxidase-streptavidin conjugate, and exposed to 3,3'-diaminobenzidine (0.07 mg) and H2O2-urea (0.06 mg) in tap water for 10 min. Sections were counterstained with 0.3% methyl green. Nuclei with DNA fragmentation stained brown, whereas normal nuclei stained green. Negative controls were performed by omitting TdT enzyme.
Immunohistochemistry for apoptosis-related proteins. Serial
cryopreserved sections (5 µm) were used in these studies. In each instance,
a series of three sequential sections were stained for CD3 and two additional
proteins so that 
10 µm of depth separated the first from the last
slide. Sections were fixed for 20 min at 4°C in acetone, washed with PBS,
and then quenched of endogenous peroxidase activity with 3%
H2O2. After nonspecific binding was blocked for 20 min
with 1% BSA, sections were incubated for1hat room temperature with antibodies
directed against Fas (1 µg/ml; BD Transduction Laboratories, San Diego,
CA), FasL, Bax, or FLIP (4 µg/ml; Santa Cruz Biotechnology, Santa Cruz,
CA), or Bcl-2 or Bcl-xL (4 µg/ml; Neomarkers, Fremont, CA). All antibodies
were mouse monoclonal antibodies raised against the human form of the protein
with no species reactivity tested against rabbits. After the sections were
washed with PBS, they were incubated with biotinylated secondary antibody
(AutoProbe III, Biomeda) for 20 min at 37°C. Antibody-biotin conjugate was
detected with 3-amino-9-ethylcarbazole and counterstained with hematoxylin.
Negative controls were performed using species-specific IgG isotype controls
(Santa Cruz Biotechnology) instead of primary antibody.
Identical high-powered fields (x400 magnification) from each set of three slides were then coordinated to find CD3+ cells that also stained for these apoptosis proteins. A total of 12 such sections were analyzed over three allografts per postoperative time interval and treatment group.
Immunoblotting for Fas and FasL. Donor left ventricular tissue was homogenized in lysis buffer (50 mM Tris · HCl, pH 8, 150 mM sodium chloride, 1% Nonidet P-40, 0.1% SDS, and 0.5% deoxycholic acid). Aliquots (100 µg) were then denatured, separated on 12% SDS-polyacrylamide gels, and transferred to nitrocellulose membranes (Osmonics, Westborough, MA). After they were blocked with 5% nonfat milk in wash solution (Tris-buffered saline with 0.05% Tween 20), membranes were incubated overnight at 4°C with antibodies against FasL (Santa Cruz Biotechnology), Fas (BD Transduction Laboratories), pan-actin (Neomarkers), caspase-8 (BioVision, Mountain View, CA), and caspase-9 (BioVision). Membranes were then incubated for1hat room temperature with horseradish peroxidase-labeled rabbit anti-goat (KPL, Gaithersburg, MD) or goat anti-mouse (KPL) secondary antibody. Bands were visualized with Lumiglo (KPL) chemiluminescence substrate. Blots developed after omission of primary antibody were used as negative controls.
Statistical analysis. Statistics were performed using the SPSS (version 9.0) software package for Microsoft Windows (SPSS, Chicago, IL). Values are means ± SE. All comparisons between the control and gene therapy groups were done using unpaired Student's t-tests with unequal variance. P < 0.05 was considered significant. For each transplanted heart, CD3 T cell expression of apoptosis regulatory proteins (Fas, FasL, FLIP, Bax, Bcl-2, and Bcl-xL) was calculated as the number of CD3+ cells over total infiltrating CD3 cells summed across all examined fields (6–8 high-powered fields per heart). Linear regression was performed with SPSS with P values calculated by one-way ANOVA.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
As we reported previously, intragraft expression of IL-10 improved graft survival from a median of 9 days in the control group to 28 days in the gene therapy group (32, 36). This enhanced survival arose from improved acute rejection scores for the gene therapy group during each phase of the posttransplantation course compared with controls treated with "empty" liposome or lipsome-antisense IL-10 gene. By POD 4–5, control and IL-10 antisense gene-treated allografts exhibited severe rejection (scores 3–4), with myocyte necrosis and extensive lymphocytic infiltrates, whereas gene-treated grafts have, on average, only perivascular infiltrates (scores 1–2; Fig. 1D). Graft failure rapidly ensued after this degree of severe rejection.
Graft-infiltrating T lymphocyte loss through apoptosis. IL-10 gene therapy significantly reduced the number of CD3+ T lymphocytes within the allograft from 42.9 ± 2.9 to 28 ± 2.0% (P < 0.05). CD4+ (from 28.7 to 19.2%) and CD8+ (from 19.8 to 12.5%) subpopulations were diminished in similar proportions. To further define the effect of intragraft IL-10 expression on this quantitative change in T cell populations, the postoperative period was divided into an early phase with low IL-10 expression (POD 1–2), a second phase with rising IL-10 expression (POD 4–5), and a third phase with constant peak IL-10 expression (POD 7–8).
During the early phase, there was no significant difference between gene therapy and control allografts with regard to CD3+, CD4+, or CD8+ cells undergoing apoptosis as measured by TUNEL (Table 1). However, as intragraft IL-10 expression rose during the second and third postoperative phases, a significantly higher percentage of CD3+ T cells was positive by TUNEL in the gene therapy group than in the control group (P < 0.01). The increase in T lymphocyte apoptosis with IL-10 gene therapy was not limited to a single subpopulation, because CD4+ and CD8+ cells showed similar trends. The percentage of TUNEL-positive CD4+ and CD8+ cells increased seven- and sixfold, respectively, in the gene therapy group compared with the control group on POD 3–6. The difference in the apoptosis of CD4+ and CD8+ T lymphocyte subpopulations between controls and gene-treated grafts persisted throughout the peak phase of IL-10 expression (POD 7–8 and 10–11).
|
CD3+ T cell expression of Fas. As shown in Fig. 2A, IL-10 gene therapy of cardiac allografts increased Fas expression on CD3+ T cells. Although no difference existed in the early phase, more T cells expressed Fas in gene-treated allografts than among controls during the period of rising IL-10 expression (69.5 ± 6.8 vs. 3.2 ± 0.7%, P < 0.01; Fig. 2, A and B). However, as CD3+ T cells were eliminated over this time course, a smaller percentage of CD3+ T cells remaining on POD 7–8 expressed Fas (25.1 ± 9.5%), although this still exceeded time-matched controls (3.8 ± 1.2%, P < 0.05). IL-10 gene therapy did not influence expression of FasL by T cells at any point during the postoperative course. The higher Fas and unaltered FasL expression within donor heart tissue after IL-10 gene therapy was corroborated by immunblotting of myocardial tissue lysates (Fig. 2C).
|
Activation of caspase-8. Caspase-8 (FLICE) forms a direct link between the activation of Fas (CD95) and the caspase pathway (4, 21, 28). Caspase-8 is a 55-kDa cytosolic protein that is synthesized as an inactive proenzyme. Activation of caspase-8 involves a two-step proteolysis: the cleavage of caspase-8 to generate a 43- and a 12-kDa fragment, which is further processed to 10 kDa. The p43 is then cleaved to yield p26 and the release of the activation site containing p18. To assess activation of caspase-3, total heart tissue lysates from the left ventricle were assessed by immunoblotting with an antibody that recognizes the full-length 55-kDa form and the active 43-kDa subunit.
Caspase-8 is activated significantly in the myocardium gene group, but not in controls (Fig. 3A). The fraction of activated caspase-8 was only 2.6% in the control group at POD 1–2 and slightly increased in the later stage. The fraction of activated caspase-8 was sevenfold higher in the IL-10 gene therapy group than in the control group at POD 4–5. At POD 7–8, the amount of activated caspase-8 was slightly reduced in the gene group compared with POD 4–5 but was still 2.3-fold higher than in the control group at the same time.
|
CD3+ T cell expression of FLIP. In addition to enhancing CD3+ T cell expression of proapoptotic molecules, Fas, IL-10 gene therapy of cardiac allografts also suppressed expression of antiapoptotic molecules in these cells. Expression of FLIP (Fig. 3B) increased among graft-infiltrating CD3+ T cells from 30.3 ± 5.3% at POD 1–2 to 45.6 ± 4.1% by POD 4–5 and 49.7 ± 7.7% by POD 7–8 in control allografts. Ex vivo IL-10 gene therapy markedly reduced FLIP expression among CD3+ T cells. As early as POD 1–2, IL-10 gene therapy ablated FLIP expression in T cells, and reconstitution of this expression was incomplete at later postoperative time points (P < 0.01 for differences in FLIP expression in all 3 postoperative phases).
CD3+ T cell expression of Bax. In addition to increasing Fas expression in graft-infiltrating T cells, IL-10 gene therapy also upregulated proapoptotic Bax expression during POD 4–5 compared with controls (Fig. 4A). In the gene therapy group, 46.2 ± 6.9% of graft-infiltrating CD3+ T cells expressed Bax compared with 1.5 ± 0.7% in control animals (P < 0.01). This difference is lost by POD 7–8 as the percentage of Bax-expressing CD3+ cells diminishes in the gene-treated group and increases in the control group (25.1 vs. 24.3%, P = not significant). No staining of sections was seen with substitution of isotype control antibodies for antibodies directed against Bax or Fas.
|
CD3+ T cell expression of Bcl-xL. The effect of IL-10 gene therapy on Bcl-xL expression in CD3+ T cells was similar. As shown in Fig. 4B, IL-10 gene therapy virtually eliminated expression of Bcl-xL in the early postoperative period compared with controls (22.5 ± 9.0 vs. 0.1 ± 0.1%, P = 0.03). Bcl-xL expression rises slowly over POD 4–5 (4.4 ± 2.0%) and 7–8 (5.5 ± 2.9%) among CD3+ T cells from gene-treated allografts, but, as with FLIP, levels of expression comparable to controls were never attained. By POD 7–8, more CD3+ T cells from genetreated grafts continued to express Bcl-xL than T cells from controls (22.1 ± 6.1 vs. 5.8 ± 3.9%, P = 0.04). Interestingly, immunohistochemistry for Bcl-2 showed no significant differences in expression among CD3+ T cells from IL-10 gene-treated and control animals at any postoperative time. Again, no staining was seen in the immunhistochemistry of these sections with an isotype control antibody substituted for the primary.
Activation of caspase-9. Mitochondrial proapoptotic protein, such as the Bcl-2 family, triggers the procaspase-9 (46 kDa), binds ATP and cytochrome c, and subsequently proteolytically cleaves itself to form the active caspase-9 (35 kDa) (4, 21, 38). Activated caspase-9 was seen in control and IL-10 antisense and IL-10 sense gene-treated allografts (Fig. 5A). However, the fraction of activated caspase-9 was more than two-fold higher in IL-10 sense gene-treated allograft than in empty lipsome- and IL-10 antisense gene-treated allograft in all time periods.
|
Fas-mediated T cell apoptosis and rejection. CD3+ cells
that stain by the TUNEL method have higher coincident staining for Fas in
IL-10 gene-treated grafts than controls. During POD 4–5 and
7–8, 
50% of these TUNEL-positive CD3 cells coexpress Fas
in the gene-treated grafts compared with <15% of controls (P <
0.01 for both postoperative intervals; Fig.
5B). In contrast, expression of Bax in TUNEL-positive CD3
cells doubles after gene therapy of grafts (35 vs. 16.7%, P <
0.01), but only during peak IL-10 expression over POD 4–5
(Fig. 5B). This
difference disappears rapidly, so that by POD 7–8, no
difference exists between the extent of Bax expression (42.7 vs. 46.8%,
P = not significant).
The importance of the sustained Fas overexpression in CD3 cells induced by IL-10 gene transfer is emphasized by the relations diagrammed in Fig. 6A. Among grafts undergoing acute cellular rejection, a more prevalent expression of Fas among graft-infiltrating CD3 cells correlates to a higher rate of apoptosis measured by the TUNEL method (r = 0.84, P = 0.02) (22). This contrasts to the strong inverse relation shown in Fig. 6 between the rate of apoptosis and expression of the cellular inhibitor of Fas signaling, FLIP, among CD3+ cells (r = –0.87, P < 0.01). The ratio of the fraction of CD3 cells expressing Fas to the fraction expressing FLIP exhibits a significant correlation to the histological rejection score (Fig. 6B; r = –0.86, P < 0.01). As the fraction of CD3 cells with Fas expression increases relative to cells with FLIP expression, rejection scores improve. Neither Bax nor Bcl-xL shows significant correlation to the rate of apoptosis by TUNEL or to histological rejection grade (data not shown).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
The ability of IL-10 overexpression in this model to increase apoptosis of T cells diverges from the reported antiapoptotic effect of this cytokine in vitro. Under the conditions of IL-2 withdrawal and deprivation, IL-10 treatment rescues activated T cells from apoptosis (28, 33). Additionally, when stimulated toward AICD, preactivated T cells from normal human volunteers do not exhibit augmented apoptosis in response to exogenous IL-10 administration (13). However, two other reported findings confound this apparent neutral or antiapoptotic effect of IL-10 on apoptosis. 1) Although exogenous IL-10 does not normally increase apoptosis, blockade of endogenous IL-10 through monoclonal antibodies reduces AICD among preactivated T cells (5, 13). 2) In peripheral blood mononuclear cells or isolated Th1 cell lines, IL-10 treatment does increase apoptosis (5, 19). The data in our study come from an in vivo model of alloreactivity, which has important differences from prior in vitro systems used to investigate the effect of IL-10 on T cells. These differences include a time-dependent expression of IL-10, TCR stimulation through alloantigens, rather than cross-linking by antibodies, and, most importantly, the simultaneous presence of cellular elements other than T cells in the local environment. As shown by the effect of IL-10 on peripheral blood mononuclear cells (19), the presence of additional cellular elements can change the neutral effect of IL-10 to a proapoptotic effect, suggesting that interactions between T cells and other cell types can modulate the influence of IL-10.
The complexity of this in vivo model is reflected by the variety of cellular mechanisms of apoptosis modulated in response to IL-10 gene treatment. Whereas AICD within in vitro models is triggered by upregulation of Fas/FasL with subsequent autocrine and paracrine induction of apoptosis (3, 16), additional pathways contribute to T cell apoptosis after allograft IL-10 gene therapy. Although Fas expression on T cells is increased by IL-10, a proapoptotic shift in the balance of Bcl-2 family proteins also occurs, with greater Bax and less Bcl-xL expression. The significance of these changes is unclear because alterations in Bax expression are temporary, and neither Bax nor Bcl-xL expression correlates to graft rejection or the rate of infiltrating T cell apoptosis. Alterations in this mitochondrion-based apoptosis pathway could simply reflect alterations in cytokine secretion profiles. A mechanistic dichotomy exists between AICD and cytokine deprivation and is supported by experiments from transgenic mice. T cells from Bcl-2 transgenic mice are protected from apoptosis associated with IL-2 withdrawal, but not from AICD (11). Conversely, T cells from lpr/lpr Fas knockout mice exhibit defective AICD but can undergo apoptosis normally in response to cytokine deprivation (8, 11). Thus the changes in Bcl-2 family proteins after IL-10 gene therapy could reflect differences in cytokine growth factor availability, perhaps because of the suppressive effect of IL-10 on release of Th1-derived cytokine growth factors such as IL-2 (30). The ability of other cytokines to substitute for IL-2 could account for the transience in these changes in Bax expression.
Elevated Fas expression on T cells of the IL-10 genetreated allografts sensitizes them to AICD. The downregulation of FLIP expression by IL-10 removes a blockade to procaspase-8 activation by Fas and its cytoplasmic adapter molecules (23). These two effects quantitatively augment and functionally improve signaling through the Fas-FasL pathway of graft-infiltrating T cells after IL-10 treatment. In T cells activated in vitro with anti-CD3 antibodies, addition of IL-10 diminishes Fas expression, although this does not influence the sensitivity of these cells to AICD (33). However, antibodies to IL-10 blunt increases in Fas mRNA expression caused by nonspecific T cell stimulation with concavalin A among lymphocytes isolated from spleens of septic mice (5). Again, IL-10 has opposite effects in in vivo models compared with results from in vitro systems, supporting the importance of neighboring cell types in determining T cell responses. In regard to FLIP, costimulation through CD28 is known to increase its expression and expression of Bcl-xL (6, 26). IL-10 attenuates costimulation through reduction of APC B7 expression and direct effects on CD28 tyrosine phosphorylation (2, 31). Thus, although appropriately costimulated T cells counteract greater Fas and FasL expression through FLIP blockade of caspase-8 activation, T cells without costimulation have less ability to inhibit this downstream caspase activation and are more prone to AICD.
The functional consequences of these altered apoptosis mechanisms are reflected by the coordinated reduction in graft-infiltrating T cells with improved histology and graft survival after IL-10 gene therapy. The preeminence of the Fas signaling pathway for achieving these effects is emphasized by the correlation between the fraction of Fas and FLIP expression to the rate of T cell apoptosis and to rejection score. The improvements in histological rejection and graft survival are widely supported by other investigators who have used IL-10 gene therapy in animal transplant models (9, 15, 35). Although most investigators have used the viral IL-10 in these studies (9, 15, 35), cellular IL-10 gene transfer has also led to successful local intragraft immunosuppression (14, 37). Additionally, human IL-10 specifically acts to suppress APC from rabbits (17, 29), indicating cross-species reactivity. However, despite these effects, immunologic tolerance is not achieved with these transplant models. Data from our studies suggest multiple reasons for this failure. First, Fas and Bax expression decline from POD 3–6 to 7–10 in T cells from gene-treated animals. Additionally, FLIP and Bcl-xL expression slowly increase over the posttransplant time course in the same cells. These trends indicate that T cell sensitivity to apoptosis through AICD and cytokine deprivation declines over time after IL-10 gene therapy. These results could indicate selection of a relatively apoptosis-resistant T cell population through early apoptosis or activation of alternative routes of costimulation (intercellular adhesion molecule and inducible costimulator) later in the postoperative course after gene therapy.
In conclusion, this study demonstrates that IL-10 overexpression within an in vivo model of cardiac allograft transplantation augments apoptosis of infiltrating T lymphocytes in the CD4 and CD8 subpopulations. Enhanced apoptosis of T cells within this model reflects greater sensitivity to AICD through upregulation of Fas and downregulation of FLIP, with cytokine deprivation through higher Bax and lower Bcl-xL less important. However, these effects are temporary and begin to diminish by POD 7–10, leading to eventual graft failure. The variation in these results from studies of IL-10 in vitro demonstrates the critical role of other cells in the local environment in T cell behavior and points out the limitations of cell culture-based models for evaluating cytokine behavior within the organism.
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Oshima, G. Cui, T. Tung, O. Okotie, H. Laks, and L. Sen Exogenous IL-10 overexpression reduces perforin production by activated allogenic CD8+ cells and prolongs cardiac allograft survival Am J Physiol Heart Circ Physiol, January 1, 2007; 292(1): H277 - H284. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Miyamoto, T. Kaneko, M. Yamashita, Y. Tenda, M. Inami, A. Suzuki, S. Ishii, M. Kimura, K. Hashimoto, H. Shimada, et al. Prolonged skin allograft survival by IL-10 gene-introduced CD4 T cell administration Int. Immunol., June 1, 2005; 17(6): 759 - 768. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
5 samples at each postoperative interval. *P < 0.01.
