| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 From the Department of Ophthalmology, Osaka University Medical School, Osaka, Japan; 2 Department of Ophthalmology, Sir Run Run Shaw Hospital, Zhejiang Medical University, Hangzhou, China; 3 Kobe Institute of Health, Kobe, Japan; and the 4 Department of Ophthalmology, Ehime University, School of Medicine, Ehime, Japan.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
METHODS. C3H/He mouse corneal lenticules were grafted around the limbus in BALB/c mice, and alloepithelial rejection was assessed by microscopic evaluation. The relation between rejection scores and DTH or CTL responses to donor antigens was assessed by Spearman correlation analysis. Suppression of DTH responses by induction of anterior chamber–associated immune deviation (ACAID) was used to evaluate the contribution of DTH responses to allograft rejection. CTL responses were evaluated by in vitro and in vivo depletion of CD4+ or CD8+ cells.
RESULTS. DTH responses, which developed 2 weeks postoperatively,
correlated significantly with rejection scores (correlation coefficient
r = 0.55). ACAID induction by anterior chamber
inoculation of C3H/He splenocytes significantly suppressed allospecific
DTH responses and alloepithelial rejection. While allospecific CTL
responses also developed 2 weeks postoperatively and increased by 4
weeks, CTL responses did not exhibit positive correlation with
rejection scores (r = -0.36, P = 0.076). The CTL responses were mediated exclusively by
CD8+ cells. Although in vivo depletion of CD8+
cells abolished the induction of CTL responses, it did not prevent
allograft rejection. Immunohistochemistry showed infiltration of
CD4+ and CD8+ cells into the alloepithelium.
Positive staining for interferon 
but not interleukin 4 further
implicated the participation of a DTH response.
CONCLUSIONS. Allograft rejection after keratoepithelioplasty appears to be mediated primarily by DTH responses, with CTL responses playing only a minor role perhaps in modifying the development of rejection.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Allograft rejection in corneal transplantation is believed to develop via a T-cell–mediated immune response. Specifically, delayed-type hypersensitivity (DTH) appears to play a dominant role in corneal allograft rejection as shown by several studies in the murine model of penetrating keratoplasty (PKP).6 7 8 Moreover, Joo and colleagues7 reported no significant difference in activity of cytotoxic T lymphocytes (CTL) between mice with accepted versus rejected corneas, suggesting that CTL responses may not play an important role in allograft rejection. > Yet, other studies report that allospecific CTLs do develop in rats with rejected corneas, but not in rats with accepted corneas.7 10 Therefore, it remains unclear as to whether both DTH and CTL immune effector mechanisms participate in allograft rejection in rodent orthotopic PKP models.
However, the immune effector mechanisms involved in KEP likely differ from PKP for several reasons. First, rejection in KEP occurs mainly against the corneal epithelium, and does not occur against the endothelium. Second, grafts are secured around the limbus, which has abundant Langerhans cells allowing for greater host recognition of the grafts. Third, since graft components are separated from the recipient’s anterior chamber, anterior chamber-associated immune deviation (ACAID) is not induced by the surgery itself. In contrast, in PKP, the DTH response is impaired in recipients with accepted corneal allografts, and ACAID is induced by the grafted corneal button itself.6
The murine KEP model we have developed is useful for investigating the role of DTH in orthotopic corneal transplantation, because it can directly examine the contribution of the DTH response without the overlaying influence of ACAID. 5 We previously reported that anterior chamber priming by alloantigens before grafting induced suppression of the DTH response and reduction of alloepithelial rejection in this murine KEP model with minor disparate grafting.11 This suggested that DTH participates in alloepithelial rejection after KEP, although the role of other effector mechanisms was not examined. In the present study, we evaluated the contribution of both DTH and CTL responses to the immunopathogenesis of alloepithelial rejection after KEP, with examination of the role of specific T cell subsets using in vitro and in vivo depletion.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Mouse KEP and Clinical Evaluation
The method of KEP grafting has been described
previously.5
In brief, full thickness cornea without
scleral tissue was collected by excision along the limbus of the donor
eye and cut into four 1.0-mm x 2.5-mm lenticules. Under
anesthesia, the recipient corneal epithelium was scraped off
completely, and adjacent limbal conjunctiva was removed
circumferentially. Three donor corneal lenticules were secured around
the recipient corneal limbus with two interrupted 11-0 nylon sutures
(Alcon Surgical, Dallas, TX). Alloepithelial rejection was evaluated
under an operating microscope and graded using previously established
criteria, based on degree of corneal edema or opacity (0–4) and
neovascularization (0–4). The rejection score was calculated as the
sum of the edema/opacity score and the neovascularization score
(0–8). Eyes with technical failures (such as anterior chamber
perforation, hyphema, cataract, or postoperative infection) were
excluded from the study.
Assay for DTH Reaction and Induction of Anterior
Chamber–Associated Immune Deviation
At the indicated days after KEP, right ears were challenged using
intrapinnal injection of C3H/He splenocytes (1.5 x
106 cells in 20 µl phosphate-buffered saline
[PBS]) that had been lethally irradiated by cesium 137 (3000 cGy).
PBS was injected into control left ears. Ear swelling was measured at
48 hours after antigen challenge using a micrometer (Mitutoyo, Tokyo,
Japan). DTH responses were expressed as specific ear swelling =
(48-hour measurement of right ear - 0-hour measurement of right
ear) - (48-hour measurement of left ear - 0-hour
measurement of left ear).
ACAID was induced by inoculation into the left anterior chamber of allogeneic C3H/He splenocytes (3.5 x 105 cells in 5 µl PBS) that had been lethally irradiated (3000 cGy). KEP was performed on the right eye 7 days after ACAID induction. KEP recipients inoculated with irradiated syngeneic BALB/c splenocytes served as negative controls.
Cell-Mediated Cytotoxicity
Spleens removed from BALB/c mice were processed into single
cell suspensions and resuspended at a concentration of 6 x
106 cells/ml in RPMI-1640 medium supplemented
with 10% fetal bovine serum, 5 x 10-5 M
2-mercaptoethanol, 10 mM HEPES, and 0.1 mM nonessential amino acids.
These cells were stimulated with irradiated C3H/He spleen cells (6 x 106 cells/ml) in vitro for 5 days. After this
in vitro boosting, 100-µl aliquots of the splenocyte effector cell
suspensions were added to round-bottomed microtiter plates at various
concentrations. 51Chromium-labeled target cells
were added to the wells in triplicate to produce effector-to-target
ratios (E/T ratios) ranging from 100/1 to 12.5/1 and incubated at
37°C for 4 hours. The target cells used were RDM-4
(H-2k) for allogeneic CTLs,
P815(H-2d) for syngeneic controls, and YAC-1 for
measurement of natural killer (NK) activity. After centrifugation, 100
µl of each supernatant was collected, and counts per minute (cpm)
were measured by a gamma counter. Specific cytotoxicity of each sample
was calculated by the following formula: % specific lysis =
[(experimental cpm - spontaneous cpm)/(maximum release cpm - spontaneous cpm)]x 100. Maximum 51Cr release
was determined by measuring the released amount into the medium by
51Cr-labeled target cells incubated with 2%
Triton X-100.
In Vitro Depletion of T-Cell Subsets
Before specific cytotoxic assay, splenocyte suspensions stimulated
in vitro by irradiated C3H/He spleen cells for 5 days were incubated in
diluted anti-L3T4 ascites fluid (50x), anti–Lyt-2.2 antibody (500x;
Cedarlane, Hornby, Ontario, Canada), or anti–asialo GM1 antibody
(50x; Wako, Osaka, Japan) on ice for 1 hour. Splenocytes were then
resuspended in a 1:10 ratio with diluted rabbit complement (Cedarlane)
and incubated for 1 hour at 37°C. Depletion of specific T-cell
subsets was confirmed by FACS analysis (Becton Dickinson, Bedford, MA).
Direct fluorescein staining was used to detect
L3T4+ or Lyt-2+ cells using
fluorescein isothiocyanate (FITC)–conjugated rat anti–mouse CD4
antibody (Serotec, Oxford, UK) or FITC-conjugated rat
anti–mouse Lyt-2 antibody (Cedarlane). The depleted effector cells
were used for the cytotoxic assay.
Preparation of Monoclonal Antibodies and In Vivo Depletion of
T-Cell Subsets
The GK 1.5 hybridoma cell line (anti-L3T4) and the 53.6.72
hybridoma cell line (anti–Lyt-2) were propagated in nude mice, with
monoclonal antibodies isolated from ascites fluids. Antibodies were
purified using E-Z-SEP and Hi Trap Q ion exchange chromatography
(Pharmacia Biotech, Piscataway, NJ). Protein concentrations
were determined using BCA reagents (Pierce, Rockford, IL).
In vivo depletion of T-cell subsets was conducted by intraperitoneal injection of monoclonal antibodies, starting on 3 consecutive days 1 week before the KEP procedure, and continued twice a week until the end of the observation period at a dose of 0.1 mg/mouse for L3T4 antibody and 0.2 mg/mouse for Lyt-2 antibody, respectively. Specific depletion was confirmed by FACS analysis of splenocytes obtained from injected mice.
Immunohistochemistry
Eyes were enucleated at day 3, day 7, day 14, day 21, 1
month, 2 months, and 4 months after KEP. Eyes used for Thy1.2 staining
were fixed in paraformaldehyde-lysine-periodate (PLP) fixative
for 24 hours at 4°C, then embedded in paraffin, cut into
6-µm–thick sections, and deparaffinized. Other samples were prepared
in 7-µm–thick frozen sections and fixed in acetone at -20°C.
The sections were then incubated with 0.3%
H2O2 for 30 minutes to
quench endogenous peroxidase activity. Immunohistochemistry was
performed using a streptavidin–horseradish peroxidase system
(Kirkegaard & Perry Laboratories, Gaithersburg, MD) according
to the manufacturer’s instructions. The primary antibodies used were
rat anti–mouse Thy1.2 (Becton Dickinson), rat anti–mouse L3T4
(GK1.5), rat anti–mouse Lyt-2 (53.6.72), rat anti–mouse interferon
(RMMG-1), and rat anti–mouse interleukin (IL)-4 (Pharmingen, San
Diego, CA). Sections were developed with diaminobenzidine
tetrahydrochloride (Vector Laboratories, Burlingame, CA) and
counterstained with methyl green.
Statistical Analysis
Data were analyzed using the two-tailed Student’s
t-test or Mann–Whitney U test as appropriate, and
differences were considered significant for P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To determine the role of the DTH response in this rejection, we investigated whether the pattern of DTH induction correlates with graft rejection. DTH responses were measured at weekly intervals after KEP. The kinetics of the DTH response against alloantigen from the C3H/He donor is shown in Figure 1 A. An alloantigen-specific DTH response initially appeared at 1 week (P = 0.12) and peaked at approximately 2 weeks (P < 0.05). This DTH response was found to persist at least through s postoperative week 4 (P < 0.0005). To estimate the contribution of DTH to allograft rejection, Spearman correlation analysis was used to analyze the data at 2 weeks (Fig. 1B) . Specific ear swelling of the grafted mice were found to correlate significantly with rejection scores (r = 0.55, P < 0.05).
|
|
Furthermore, involvement of NK activity was evaluated using YAC-1 target cells. Natural killer activity was detected 2 weeks postoperatively; however, effector cells amplified 4 weeks after grafting exhibited significant but low lytic activity of YAC-1 target cells (% specific lysis at an E/T ratio of 50/1 was 2.6% for KEP recipients and 0.3% for naive BALB/c; n = 4 for each group). The specific lysis of YAC-1 target cells was significantly suppressed by depletion of asialo GM1–positive cells (data not shown).
To further characterize the T-cell subset with allospecific killer activity, we next measured CTL activity after in vitro depletion of specific T-cell subsets. Splenocytes from grafted BALB/c mice, depleted of CD4+ cells by in vitro incubation with anti-L3T4 antibody and rabbit complement, were assayed for allospecific killer activity. Although this treatment eliminated 92% of L3T4+ cells, CTL activity to lyse RDM-4 target cells was not significantly affected (Fig. 3 A). Next, splenocytes from grafted BALB/c mice, depleted of CD8+ cells by incubation with anti–Lyt-2 antibody and complement, were also assayed for CTL activity. This treatment was shown to eliminate 96% of CD8+ cells and was found to abolish allospecific CTL activity (Fig. 3B , P < 0.05). These findings indicate that CTL activity after KEP is mediated by the CD8+ subset rather than the CD4+ subset of T cells.
|
|
|
|
|
Finally, we investigated whether infiltrating inflammatory cells in the
KEP model produce either interferon , associated with the Th1 immune
response, or IL-4, associated with the Th2 immune response. During the
acute phase of rejection, interferon –positive cells were found to
infiltrate the lenticules and corneal epithelium (Fig. 7B)
. To the
contrary, no IL-4+ cells were detected during the
observation period.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
During the acute phase of the rejection, massive infiltration of
T cells was observed in the lenticules and donor-derived epithelium
(Fig. 7A)
. Both CD4+ and
CD8+ cells were present, the former being
relatively more abundant. Immunostaining of the alloepithelium revealed
the infiltration of interferon –positive cells, indicating that
Th-1 or Tc-1 cells had migrated into the donor-derived tissue and
mediated allograft destruction by secreting proinflammatory cytokines.
Because upregulation of DTH is mediated mainly by Th-1 cells, these
findings support the hypothesis that DTH is responsible for rejection
in our KEP model.
Using the combination of B10.D2 donors and BALB/c recipients, we previously determined that anterior chamber injection of donor specific lymphocytes reduces the incidence of alloepithelial rejection and suppresses allospecific DTH responses.11 This combination, disparate only in multiple minor histocompatibility antigens, induces obvious alloepithelial rejection. In the present study, we used a donor and recipient combination disparate in both major and multiple minor histocompatibility antigens. Although the additional disparity of major histocompatibility antigens might induce distinct antigen processing, anterior chamber injection of donor antigens still induced a reduction in clinical scores and an allospecific DTH response. Because ACAID induces suppression of DTH with preservation of normal CTL and humoral antibody responses,12 these results provide further evidence that alloepithelial rejection in KEP is mediated by DTH.
To examine the contribution of DTH to allograft rejection, we performed KEP after in vivo depletion of CD4+ cells (data not shown). Although some DTH suppression was achieved by depletion of CD4+ cells, the allospecific DTH response could not be completely abolished. Because DTH developed even with 95% depletion of CD4+ cells, we could not evaluate the contribution of DTH responses to allograft rejection using this approach. Indeed, DTH responses are reported to be capable of being mediated by CD8+ and CD4+ cells.13 In the KEP model, CD8+ cells may be inducing a DTH response, or enhancing a DTH response cooperatively with other effector cell subsets, when CD4+ cell counts are low. To definitively demonstrate the contribution of DTH responses to allograft rejection using this approach, it would be necessary to use an experimental setting in which DTH responses are significantly inhibited or abolished.
In vitro depletion of CD8+ cells demonstrated that the cytolytic activity of activated CTL precursors measured in our system was mediated by CD8+ cells (Fig. 3) , whereas cytolysis by CD4+ T cells or NK cells was a minor event (Fig. 3A) . Furthermore, in vivo depletion of CD8+ cells abolished CTL responses (Fig. 6A) , indicating that the induction phase of CTL activity is also mediated by CD8+ cells. Abolishment of the CTL response did not suppress allograft rejection (Fig. 6B) . These results further support the notion that the CD8+ cell–mediated CTL response does not play an important role in allograft rejection.
Interestingly, suppression of the CTL response caused an increase
in rejection scores after the acute phase of rejection (Fig. 6B)
, with
rejection scores and CTL responses showing an inverse correlation (Fig. 4)
. This suggests that development of CTL responses might have an
inhibitory effect on alloepithelial rejection. The animals that
developed strong CTL responses may have concomitantly acquired
suppressor CD8+ subsets, which in turn regulated
DTH responses by CD4+ subsets. Depletion of
CD8+ cells might impair this suppressor activity,
and thus contribute to increased rejection scores. Another possible
explanation of this phenomenon may be the unique kinetics of the
grafted epithelium. In the KEP model, the denuded host cornea is
resurfaced by donor-derived epithelium within a few days after surgery;
however, subsequent epithelial rejection results in destruction of the
donor-derived epithelium and centripetal movement of host-derived
epithelium. Consequently, after the onset of rejection, the host
corneal bed is surfaced by a mosaic-like mixture of both recipient- and
donor-derived epithelium (unpublished observations). When activated
allospecific CTLs lyse donor-derived epithelium in the early phases of
rejection, a decrease in alloantigen density may develop. Although this
destruction of donor-derived epithelium by CTLs would itself constitute
rejection, antigen-specific CTLs attack individual donor-derived
epithelial cells in a pinpoint fashion as Ando et al.14
demonstrated in the CTL-mediated liver disease model. In the PKP model,
the attacked corneal endothelium suffers a devastating blow that leads
to graft failure. In contrast, in KEP grafting CTLs will not affect the
endothelium, only the corneal epithelium, the latter of which has a
strong ability to proliferate. Thus, CTL-induced inflammation may not
have a fatal impact on KEP grafting. The resultant decrease in antigen
load, in turn, might alter or diminish enhancement of allospecific DTH.
To the contrary, when a substantial amount of donor-derived epithelium
remains as a result of insufficient CTL activity, the major
histocompatibility complex–restricted stimulation of Th1 appears to
elicit sufficient interferon release to induce a DTH
response.15
Moreover, Yamada and Streilein16
suggest that the corneal epithelium on the graft itself promotes the
development of DTH by secretion of inflammatory cytokines. Taken
together, these observations suggest that the CTL response does not
play a major role in rejection in KEP, but rather that it modifies
development of alloepithelial rejection.
In contrast to PKP, KEP is associated with relatively more frequent and more vigorous allograft rejection. In our previous report, using C57BL/6 donors and BALB/c recipients disparate in both major and minor histocompatibility antigens, rejection occurred in 100% of animals after KEP.5 In a PKP model, C57BL/6 donors exhibited a 33% rejection rate.17 When a high risk group of PKP recipients were grafted with C57BL/6 corneal buttons, the rejection rate increased to 100%, whereas induction of ACAID by intravenous injection of allogeneic peritoneal exudate cells cultured with transforming growth factor-ß decreased the rejection rate to 37.5%.18 In the present study, C3H/He donors with major and minor histocompatibility disparity to recipients also produced a 100% rate of rejection. However, this high incidence of rejection was not reduced by induction of ACAID, although clinical symptoms were significantly decreased. The difference in rejection rates between these PKP and KEP models cannot be explained by the intervention of ACAID alone.
In the KEP model, donor cornea with accompanying limbal tissue is secured along the limbus of the recipient cornea, such that recipient antigen-presenting cells have ready access to foreign antigens and can trigger immunologic sensitization. In addition, host-derived T cells, which are also in close proximity to the graft tissue, can infiltrate the graft and mediate allograft rejection. Donor-derived antigen-presenting cells might also migrate into the host tissue directly and initiate allograft rejection cooperatively, because the grafted lenticules include limbal as well as central areas of donor cornea. The direct pathway of alloantigen recognition might facilitate upregulation of the immunologic response in KEP. Involvement of direct alloantigen recognition is supported by our previous finding that both major and minor antigens are related to corneal epithelial rejection in KEP.5 In a PKP model, Sano and colleagues17 speculated that allorecognition by recipient T cells must occur by the indirect pathway of alloantigen processing because of the of the paucity of bone marrow–derived dendritic cells. In this situation, minor histocompatibility antigens may compete favorably with major histocompatibility complex antigens for presentation by recipient antigen-presenting cells.
In PKP, the CTL response is presumed to be more important than DTH because the main target of rejection is the endothelium, which has a slower rate of growth than the epithelium. Direct attack of individual cells by CTLs may thus be more fatal for the endothelium than for the epithelium. This notion is supported by data from Ksander and colleagues19 who reported that donor-specific CTLs were not detected in draining lymph nodes of normal PKP recipients, but were detected in high-risk PKP recipients, and that their appearance coincided with graft rejection. In contrast, DTH responses in PKP are most likely weaker than those in KEP, because ACAID is induced by the PKP procedure itself.6
As was shown above, compared with the CTL response, the DTH response was more dominant in the immunopathogenesis of KEP. Although there are differences between KEP and PKP mechanisms that should be considered, the dominance of the DTH response over the CTL response in keratoplasty appears to also apply equally to PKP. He and associates20 reported that in vivo CD4 depletion promotes graft survival but that CD8 depletion does not in the murine PKP model. Joo and colleagues7 also reported that corneal allograft rejection after PKP of C57BL/6 donors to BALB/c mice is associated with the ability to generate strong DTH, and not CTL, responses. Thus, our KEP model is useful for increasing our understanding of the immunologic mechanisms of corneal transplantation as well as for analyzing the therapeutic effects and epithelial characteristics of clinical KEP in humans.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Submitted for publication November 4, 1998; revised April 9, 1999; accepted June 4, 1999.
Commercial relationships policy: N.
Corresponding author: Yoshitsugu Inoue, Department of Ophthalmology, Osaka University Medical School, 2-2 Yamadaoka, Suita, Osaka, 565-0871, Japan. E-mail: yinoue{at}ophthal.med.osaka-u.ac.jp
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Maruyama, J. Yamada, Y. Sano, and S. Kinoshita Th2-Biased Immune System Promotion of Allogeneic Corneal Epithelial Cell Survival after Orthotopic Limbal Transplantation Invest. Ophthalmol. Vis. Sci., November 1, 2003; 44(11): 4736 - 4741. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
