Rat Corneal Allograft Survival Prolonged by the Superantigen Staphylococcal Enterotoxin B

Zhiqiang Pan,1 Yu Chen,2 Wenhua Zhang,1 Ying Jie,1 Na Li,3 and Yuying Wu1

PURPOSE. The purpose of this study was to determine the The term superantigen (SAg) is used to describe those optimal conditions for prolonging corneal allograft survival by microbial products that activate a large portion of the T-cell inducing anergy with the superantigen staphylococcal entero- population (5%–30%), whereas conventional antigens stimu- toxin B (SEB). late only 0.01%. Superantigens differ from conventional anti- METHODS. A rat model of penetrating keratoplasty, whereby gens in that they bind to the outside of the peptide-binding Fisher344 donor corneas are implanted into Lewis recipients, groove of MHC class, thus exerting their effects as an intact was used to evaluate the effects of SEB on inhibiting immune- molecule without being processed. Furthermore, recognition mediated allograft rejection. To induce anergy, SEB was in- of SAgs by the T-cell receptor (TCR) depends only on the jected into the peribulbar space of Lewis rats. Furthermore, variable region of the TCR ␤ chain (V␤). Therefore, SAgs histopathology and immunofluorescent staining were used to stimulate both antigen-presenting cells (APCs) and T lympho- examine the levels of infiltrating CD4ϩ and CD8ϩ T lymphocytes, which leads to the massive production of cytokines, cytes and NK1.1ϩ lymphocytes. enhanced expression and/or activation of cell adhesion mole- ␮ cules, T-cell proliferation, activation-induced apoptosis, and RESULTS. By administering SEB, at doses of 90 or 120 g/kg 7 4 days before and after keratoplasty, we suppressed the episode T-cell anergy. A study has shown that injecting the SAg staph- of corneal graft rejection for a median of 12 and 30 days, ylococcal enterotoxin B (SEB) into mice produces transient, rapid hyperactivation and proliferation of T cells, which are respectively. In contrast, rejection was observed when 30 or 5 60 ␮g/kg of SEB was administered. After SEB injections, lym- eliminated by activation-induced cell death within 48 hours. phocyte infiltration into the corneal grafts was reduced, and In addition, the remaining SAg-reactive cells fail to proliferate the expression of NK1.1ϩ lymphocytes was enhanced, suggest- in response to a secondary SAg challenge. The proliferative ing that anergy may be occurring. Also, there were no differ- unresponsiveness of the secondary SAg responder T cells has ences in the number of infiltrating CD4ϩ and CD8ϩ T lympho- been termed anergy. However, on subsequent analysis, secondary SAg responder T cells may not be truly anergic because cytes cells between the control group and groups injected with 6 30 and 120 ␮g/kg SEB on postoperative days 10 and 30. they could be reacting to the second SAg exposure. Wang et al.7 showed that CD8ϩ regulatory suppressive T cells could CONCLUSIONS. Inducing anergy with the superantigen SEB pro- enforce anergy by inhibiting cell division of preactivated T longed corneal graft survival in a rat model of penetrating cells, not by the SAg response of naı¨ve T cells. keratoplasty. Therefore, these results support the possibility of Damage to healthy, transplanted tissue can be curtailed by prolonging corneal allograft survival in a clinical setting by reducing the inflammatory response of the immune system. preventing immune-mediated rejection through the administra- Our studies have shown that the bacterial superantigen SEB tion of the superantigen SEB. (Invest Ophthalmol Vis Sci. 2003; can inhibit the rejection of transplanted mouse bone marrow 44:3346–3351) DOI:10.1167/iovs.02-0845 cells and peripheral lymphocytes, suggesting that injections of SEB may induce peripheral anergy to allogeneic organ grafts. In orneal transplantation is the most common and successful ϩ addition, CD4 T cells appear responsible for maintaining this Cform of solid tissue transplantation. Although the rejection ϩ 8 rate for routine keratoplasty is only 10% in the first year, there anergy rather than the CD8 T cells. Our studies confirmed are a significant number of corneal graft failures in high-risk that the induction of transplantation anergy by SEB injections cases of keratoplasty due to immune-mediated rejection.1,2 contribute to hematopoietic chimerism, defined as the coex- Corticosteroids and cyclosporin A greatly reduce the rejection istence of host and donor cells, and that mixed lymphocyte rate of corneal allografts, but high doses and/or long-term reactions (MLRs) are significantly low. This in vivo SEB-induced administration of these drugs can produce deleterious side anergy may be associated with the clone deletion of T-helper effects, such as glaucoma, cataract formation, nephrotoxicity, (Th)1 cells. The specific dosage of SEB is critical because, if the hypertension, and hepatotoxicity.3 dosage is too high or too low, then anergy may not develop in the injected mice.8,9 Several mechanisms that can induce CD4ϩ T-cell death or unresponsiveness have been identified. These mechanisms in- 1 3 From the Beijing Institute of Ophthalmology and the Depart- clude T-cell anergy, which is due to the absence of costimula- ment of Ophthalmology, Beijing TongRen Eye Center, Capital Univer- 2 tion at the time of activation, and Fas-mediated activation- sity of Medical Sciences, Beijing, China; and the Department of Im- ␥ ␣ munology, General Hospital of PLA, Beijing, China. induced cell death (AICD). Cytokines, such as IFN- , TNF- , Supported by the National Natural Foundation of China. and IL-10, can also mediate T-cell suppression; however, the Submitted for publication August 20, 2002; revised February 25 detailed mechanisms involved in the induction of these cyto- and March 18, 2003; accepted March 27, 2003. kine-regulated T-cell death pathways have not been fully char- Disclosure: Z. Pan, None; Y. Chen, None; W. Zhang, None; Y. acterized.10 Jie, None; N. Li, None; Y. Wu, None Bacterial SAgs are a large group of polypeptides that are The publication costs of this article were defrayed in part by page produced by bacterial strains, such as Staphylococcus aureus charge payment. This article must therefore be marked “advertise- and S. pyrogenes. SAgs have been implicated in the pathogen- ment” in accordance with 18 U.S.C. §1734 solely to indicate this fact. Corresponding author: Zhiqiang Pan, Beijing Institute of Ophthal- esis of toxic-shock–like syndromes in both animal models and mology, Beijing TongRen Eye Center, Capital University of Medical in humans. In particular, the in vivo immune response to the S. Sciences, 17# Hou Gou Lane, ChongNei Street, Beijing 100730, China; aureus SAg SEB represents a useful model for studying in vivo [email protected]. cytokine regulation and the phenomenon of peripheral T-cell

Downloaded from iovs.arvojournals.org on 10/02/2021 IOVS, August 2003, Vol. 44, No. 8 Enhancement of Corneal Allograft Survival by SEB 3347

unresponsiveness. After their initial clonal expansion and cy- TABLE 1. Clinical Scoring Scheme for the Severity of Corneal tokine production, SEB-reactive T cells are thought to become Graft Rejection unresponsive to further TCR stimulation.11 T-cell deletion and unresponsiveness could be induced by oral administration or Type/Score Clinical Finding intrathymic injection of low doses of SEB, which could result in Graft opacity not only gut-associated lymphoid tissue (GALT) and thymic 12,13 0 No opacity alterations but also peripheral pivotal immune alterations. 1 Slight opacity, details of iris clearly visible The SEB used in this study was produced from S-6 cocci 2 Some details of iris no longer visible ferment, which has a molecular mass of 2.84 to 2.9 kDa and an 3 Pronounced opacity, pupil still isoelectric point of 8.6, and purified by affinity chromatogra- recognizable phy. In our previous experiments, SEB selectively decreased 4 Total opacity the percent of CD4ϩ T cells and CD4ϩ T/H2Kb cells, but had Graft edema no effect on the number of CD8ϩ cells after allogenic mouse 0 No edema cell transplantation. Accordingly, the proliferative response 1 Mild edema 2 Pronounced edema with raised transplant caused by MLRs in the recipient mouse decreased signifi- 8 3 Pronounced edema with small bleb cantly. Herein, we report that treating an inbred rat model of 4 Pronounced edema with large bleb allograft keratoplasty rejection with SEB injections into the Graft neovascularization peribulbar space at preoperative and postoperative day 7 led to 0 No vessels rat immune unresponsiveness (anergy), which persisted for 15 1 Vessels appearing in the corneal bed to 20 days after only one injection. Because SEB entered the 2 Vessels appearing in the graft periphery blood and induced CD4ϩ and CD8ϩ T lymphocyte unrespon- 3 Vessels extending deeper siveness, it appears that the effect was systemic.8 In all, we 4 Vessels extending to the center used the superantigen SEB to induce anergy for the purpose of prolonging corneal allograft survival and to observe any subsequent immune reactions. bined score of all three factors were equal to or exceeded 6. Also, the mean survival time for each group was calculated. MATERIALS AND METHODS Histopathological and Animals Immunofluorescence Evaluation Adult inbred female F344 and Lew rats were purchased from the At postoperative days 10 and 30, two graft-recipient animals were Animal Institute of Chinese Medical Academy (Beijing, China) and used killed by carbon dioxide inhalation. Afterward, the eyes were enucle- as experimental subjects between 2 and 2.5 months of age. All animals ated and fixed in a buffered formalin solution (4%). For histologic were treated according to the ARVO Statement for the Use of Animals assessment, the formalin-fixed eyes were cut into 4-␮m-thick sections in Ophthalmic and Vision Research. and subjected to hematoxylin and eosin staining. For immunofluorescence evaluation, the 4-␮m-thick sections were subjected to direct Induction of Anergy immunofluorescence using the monoclonal antibodies OX-35 (Cy- Chrome anti-rat CD4, cat. 554839), OX-8 (FITC anti-rat CD8a, cat. The SAg, staphylococcal enterotoxin B (SEB; Chinese Patent No. 554856), and R-PE–conjugated mouse anti-rat CD161a (NKR-P1A; all 01103991.4) was used for inducing anergy. SEB, at doses of 30, 60, 90, from BD Biosciences, Lincoln Park, NJ), which are markers for CD4ϩ or 120 ␮g/kg were injected into the peribulbar space of each Lew and CD8ϩ T cells and NK1.1ϩ lymphocytes, respectively. inbred rat 7 days before and after keratoplasty. Thirty-six Lew rats were divided into six groups as follows: group 1 F344/Lew (control, 0.2 mL Statistical Analysis saline buffer, peribulbar); group 2 F344/Lew (SEB, 30 ␮g/kg body weight, peribulbar); group 3 F344/Lew (SEB, 60 ␮g/kg body weight, The mean survival time (MST) and the clinical scoring data were peribulbar); group 4 F344/Lew (SEB, 90 ␮g/kg body weight, peribul- compared between the various groups by means of one-way ANOVA bar); group 5 F344/Lew (SEB, 120 ␮g/kg body weight, peribulbar); and and the independent-sample t-test, on computer (SPSS for Windows, group 6 Lew/Lew (no treatment). ver. 10.0; SPSS Science, Chicago, IL).

Orthotropic Corneal Transplantation RESULTS F344 corneal grafts (3 mm diameter) were transplanted orthotopically into anesthetized Lew eyes by a procedure described by Ma et al.3 Both Clinical Evaluation the donor graft and the recipient graft bed were scored with 3- and 2.5-mm trephines. The donor graft was sewn into place with eight Transplanted grafts, which exhibited opacity due to operative interrupted 10-0 nylon (Alcon Laboratories, Fort Worth, TX) sutures. errors, such as bleeding, suture dehiscence, and/or lens opaci- To protect the transplant, a blepharorrhaphy was attached by means of fication, within the first 3 days after surgery were omitted from two interrupted sutures and remained in place for 1 day. Also, 2 mg of the statistics (one rat in groups 2, 4, and 5 and two rats in tobramycin and 0.2 mg of dexamethasone were subconjunctival in- group 3). jected immediately after surgery and for 3 days after surgery. On postoperative day 10, all grafts from groups 1 to 3 exhibited rejection episodes (Figs. 1A–C). The grafts exhibited severe edema and infiltration into the epithelium and stroma, Assessment of Graft Survival and newly formed vessels began to penetrate the transplanted After surgery, all rats were subjected to clinical examinations by slit grafts. However, in groups 4 and 5, almost all grafts showed no lamp microscopy every day for 2 weeks followed by twice a week indications of rejection and no alteration in their normal, phys- thereafter. The transplants were evaluated using a modified form of a iological transparency (Figs. 1D, 1E). Accordingly, isografts previously described scoring system. The scoring system took into from group 6 showed no immunologic reaction (Fig. 1F). On account opacity, edema, and neovascularization14 (Table 1). An im- postoperative day 30, all grafts from groups 1 to 3 exhibited mune-mediated rejection episode was considered to occur if the com- scarring (Figs. 2A–C, respectively), several grafts from groups 4

Downloaded from iovs.arvojournals.org on 10/02/2021 3348 Pan et al. IOVS, August 2003, Vol. 44, No. 8

FIGURE 1. Clinical evaluation of rat corneal graft by slit lamp microscope at postoperative day 10. (A) Group 1, allograft control; (B) group 2, 30 ␮g/kg SEB; (C) group 3, 60 ␮g/kg SEB; (D) group 4, 90 ␮g/kg SEB; (E) group 5, 120 ␮g/kg SEB; (F) group 6, isograft control.

and 5 exhibited no immune-mediated rejection (Figs. 2D, 2E), defined as rejecting the transplant after 10 days. The mean rejec- and all grafts from group 6 remained transparent (Fig. 2F). tion index (RI) was decreased to statistically significant levels in Isografts survived for a median of more than 30 days (Table the grafts from groups 4 and 5 when compared with the control 2); however, allografts underwent clinical rejection at a median (P Ͻ 0.01). The mean opacity, edema, and neovascularization of 4 to 7 days after keratoplasty. The average transplant survival scores diminished in the allografts from groups 4 and 5 when rate of the allogenic group (group 1, F344/Lewis) was 5.17 Ϯ 1.83 compared with the control, which was statistically significant days (SD). Lew rats receiving injections of 30 ␮g/kg SEB (group (P Ͻ 0.01). The mean opacity scores for groups 4 to 6 reached 2) and 60 ␮g/kg SEB (group 3) exhibited an MST of 7 days, similar levels, but they were lower than those of groups 1 to 3 which was not a statistically significant (P Ͼ 0.05) elongation after postoperative day 30. Also, the corneal edema and neovas- of transplant survival when compared with the control (group 1). cularization scores exhibited a similar pattern (Figs. 3A–D). However, the administration of 90 ␮g/kg SEB (group 4) and 120 ␮g/kg SEB (group 5) exhibited an MST of 12 and 30 days, Histopathological and Ͻ respectively, which was a statistically significant (P 0.01) elon- Immunofluorescent Staining gation of transplant survival when compared with the control (group 1). Figure 3 shows the mean clinical scores of all groups At postoperative day 10, the grafts from the control group after graft transplantation. All the rats in groups 1 to 3 were (group 1) exhibited pronounced edema, characterized by an

FIGURE 2. Clinical evaluation of rat corneal graft by slit-lamp microscope at postoperative day 30. (A) Group 1, allograft control; (B) group 2, 30 ␮g/kg SEB; (C) group 3, 60 ␮g/kg SEB; (D) group 4, 90 ␮g/kg SEB; (E) group 5, 120 ␮g/kg SEB; (F) group 6, isograft control.

Downloaded from iovs.arvojournals.org on 10/02/2021 IOVS, August 2003, Vol. 44, No. 8 Enhancement of Corneal Allograft Survival by SEB 3349

TABLE 2. MST of Rat Graft Prolonged by Anergy-Induced SEB

MST Median Groups n (d) (Max–Min, d) Mean ؎ SD P

Allograft control 6 4, 4, 4, 4, 7, 8 4 (4–8) 5.17 Ϯ 1.83 — 30 ␮g/kg 5 5, 5, 7, 7, 8 7 (5–8) 6.40 Ϯ 1.34 0.244 60 ␮g/kg 4 6, 6, 6, 7 6 (6–7) 6.25 Ϯ 0.50 0.290 90 ␮g/kg 5 12, 12, 12, 14, 16 12 (12–16) 13.20 Ϯ 1.79 0.000 120 ␮g/kg 5 12, 21, 30, 30, 30 30 (12–30) 24.60 Ϯ 8.05 0.001 Isograft control 6 30, 30, 30, 30, 30, 30 Ͼ30 Ͼ30 —

Probabilities are treatment group results compared with the control. The MST was recorded for 30 days.

extreme thickening of the stroma, particularly in the area of revealed that there were no differences in the amount of the epithelial basal membrane. Also, pronounced mononuclear infiltrating CD4ϩ and CD8ϩ T lymphocytes between all groups infiltration was present in all layers of the corneal grafts and at postoperative days 10 and 30 (data not shown). more pronounced in the stroma and deep layers of the epithe- However, immunofluorescent staining for NK1.1ϩ lympho- lium. A dense mononuclear infiltrate and an accumulation of cytes, using the monoclonal antibody R-PE-conjugated mouse macrophages were observed at the edge of the transplant and anti-rat CD161a (NKR-P1A) showed significant differences in near the interrupted suture (Fig. 4A). the rat corneal epithelium and central stroma of the control Similar results were observed in grafts from groups 2 and 3; group and the 120 ␮g/kg SEB–injected group on postoperative however, the inflammatory infiltration was reduced slightly days 10 and 30 (Fig. 5). (Fig. 4B). In contrast, no significant infiltration of lymphocytes was observed in grafts from groups 4 and 5, especially from group 5 (Fig. 4C). However, at postoperative day 30, there DISCUSSION were no significant differences between all the transplant groups (Figs. 4D–F). Although corneal grafts have an excellent success rate com- Immunofluorescent staining for rat CD4ϩ and CD8ϩ T cells, pared with other types of transplants, 10%–20% fail as a result using the monoclonal antibodies OX35 and OX8, respectively, of immunologic rejection. In humans, high-risk keratoplasty

FIGURE 3. Average rejection score per time point for SEB-treated, negative, and isograft control grafts. There were signiﬁcant differences between the negative group and 30-␮g/kg and 60-␮g/kg SEB groups and the 90-␮g/kg and 120-␮g/kg SEB groups (P Ͻ 0.001).

Downloaded from iovs.arvojournals.org on 10/02/2021 3350 Pan et al. IOVS, August 2003, Vol. 44, No. 8

FIGURE 4. Histopathology of corneal grafts by hematoxylin-eosin staining. On postoperative day 10: (A) group 1, allograft control; (B) group 2, 30 ␮g/kg SEB; and (C) group 5, 120 ␮g/kg SEB. On postoperative day 30: (D) group 1, allograft control; (E) group 2, 30 ␮g/kg SEB; and (F) group 5, 120 ␮g/kg SEB. Mag- niﬁcation, ϫ100.

ϩ FIGURE 5. Expression of NK1.1 lymphocytes in rat transplants by im- munoﬂuorescent staining. On postoperative day 10: (A) group 1, allograft control and (B) group 5, 120 ␮g/kg SEB. On postoperative day 30: (C) group 1, allograft control and (D) group 5, 120 ␮g/kg SEB. Magniﬁca- tion, ϫ100.

Downloaded from iovs.arvojournals.org on 10/02/2021 IOVS, August 2003, Vol. 44, No. 8 Enhancement of Corneal Allograft Survival by SEB 3351

resulting in immune-mediated rejection remains the leading Effectiveness of histocompatibility matching in high-risk corneal cause of allograft failure.1,2 Besides blood type (ABO) or major transplantation. Arch Ophthalmol. 1992;110:1392–1403. histocompatibility complex (MHC) matching, immunosuppres- 2. Williams KA, Roder D, Muehlberg SM, et al. Factors predictive of sive agents such as steroids, cyclosporin A, FK506, IL-1ra, corneal graft survival. Ophthalmology. 1992;99:403–414. CTLA-4Tg, and anti-CD4 monoclonal antibodies have been 3. Ma D, Mellon J, Niederkorn JY. Oral immunization as a strategy for used for treating allograft rejection.15–20 However, more-effec- enhancing corneal allograft survival. Br J Ophthalmol. 1997;81: tive, less-toxic immunosuppressive agents are still needed to 778–784. prevent immune-mediated corneal transplant rejection. 4. Krakauer T. Immune response to staphylococcal superantigens. The mechanism of tolerance induced by SEB is poorly Immunol Res. 1999;20:163–173. understood. Immunity studies have shown that peripheral de- 5. Blackman MA, Woodland DL. In vivo effects of superantigens. Life letion in response to soluble Ag is one mechanism by which Sci. 1995;57:1717–1735. the immune system eliminates self-reactive T cells that escape 6. Rammensee HG, Kroschewski R, Frangoulis B. Cloned anergy thymic deletion. Experimental models, in which SAgs are in- induced in mature V beta 6ϩ T lymphocytes on immunizing jected into normal mice or relevant peptide Ags are injected Mls-1b mice with Mls-1a expressing cells. Nature. 1989;339:541– into TCR transgenic mice, have provided evidence for periph- 544. eral deletion.21 Another potential mechanism for silencing T 7. Wang Z-Q, Orlikowsky T, Dudhane A, et al. Staphylococcal entero- cells is functional inactivation, often referred to as anergy. toxin B-induced T-cell anergy is mediated by regulatory T cells. Immunology. 1998;94:331–339. Based largely on in vitro experiments with Th1 clones, anergy is defined as a defect in TCR-dependent proliferation that is 8. Chen Y, Peng H, Wei CC, et al. Immunotolerance and its charac- teristics induced by SEB in mice receiving allogenic cells transplan- acquired as a result of prior TCR stimulation in the absence of tation [in Chinese]. Chin J Cell Mol Immunol. 2002;18:101–103. APC-derived costimulatory signals or proliferation.22 9. Takenaka K, Fujiyama Y, Andoh A, et al. Prevention of murine Our study evaluated the potential use of SEB as a strategy for acute graft-versus-host disease by staphylococcal enterotoxin B preventing corneal graft rejection. The SEB used in our exper- treatment. Clin Exp Immunol. 2002;123:155–161. iment prolonged rat graft survival by reducing lymphocyte 10. Cauley LS, Miller EE, Yen M. Superantigen-induced CD4 T cell infiltration, edema, and neovascularization. Because graft rejec- tolerance mediated by myeloid Cells and IFN-␥. J Immunol. 2000; tion is typically due to corneal infiltration, we used corneal 165:6056–6066. opacity as a marker of immune-mediated rejection. The data in 11. Muraille E, Pajak B, Urbain J, et al. Role and regulation of IL-12 in Figure 3B show that the scores of corneal opacity in groups 4 the in vivo response to staphylococcal enterotoxin B. Int Immu- and 5 were equal to or below group two and less than groups nol. 1999;11:1403–1410. 1 and 3. Because the number of rats in each group were 12. Nishimura M, Fujiyama Y, Niwakawa M, et al. in vivo cytokine different (especially group 3) and because dexamethasone was responses in gut-associated lymphoid tissue (GALT) and spleen used in our experiments during the first three postoperative following oral administration of staphylococcal enterotoxin B. days, more experiments should be performed to verify that SEB Immunol Lett. 2002;81:77–85. plays a role in avoiding corneal allograft rejection through the 13. Goettelfinger P, Roussin R, Lecerf F, et al. T cell deletion and induction of anergy. unresponsiveness induced by intrathymic injection of staphylococ- Also, the infiltration of inflammatory lymphocytes into cal enterotoxin B. Transplant Immunol. 2002;8:39–48. grafts in group 5 (treated by SEB 120 ␮g/kg, peribulbar) was 14. Hikita N, Lopez JS, Chan C-C, et al. Use of topical FK506 in a reduced. The expression of CD4ϩ and CD8ϩ T cells was corneal graft rejection model in Lewis rats. Invest Ophthalmol Vis unchanged after SEB injection in all rat corneal grafts; however, Sci. 1997;38:901–909. there were more NK1.1ϩ lymphocytes present in the corneal 15. Nicholls SM, Williams NA. MHC matching and mechanisms of grafts from group 5 (120 ␮g/kg SEB) than in control grafts. alloactivation in corneal transplantation. Transplantation. 2001; These results suggest that T cells were unresponsive,23 but the 15;72:1491–1497. absence of an immune response must be confirmed. In addi- 16. Chan JH, Dua HS, Powell-Richards A. Effect of ABO blood group tion, more work is needed to help explain further the un- mismatching on corneal epithelial cells: an in vitro study. Br J changed expression of CD4ϩ and CD8ϩ T cells and the in- Ophthalmol. 2001;85:1104–1109. creased expression of NK1.1 T cells in the rat corneal grafts. 17. Reis A, Reinhard T, Sundmacher R, et al. A comparative investiga- In summary, these findings suggest the feasibility of using tion of FK506 and cyclosporin A in murine corneal transplantation. SEB-induced anergy as a means of reducing corneal allograft Graefes Arch Clin Exp Ophthalmol. 1998;236:785–789. rejection. Whether administration of SEB can enhance allograft 18. Fleming JC, Reid FR, Wood TO. Prevention of immune graft rejec- tolerance and permit the use of a less-intensive postoperative tion after corneal transplantation. Am J Ophthalmol. 1979;88:97– 101. treatment for patients should be verified further. The remark- 19. Lu L, Zhang WH, Sun XG. Experimental studies on the effect of the able capacity of SEB to prevent rejection in an otherwise immunosuppressant FK506 on penetrating keratoplasty rejection high-risk setting suggests that SEB may be useful as a treatment model in rat [in Chinese]. Chin J Ophthalmol. 1999;35:25–28. in high-risk keratoplasty in humans; however, well-controlled 20. Yamada J, Dana MR, Zhu SN, et al. Interleukin 1 receptor antago- clinical trials are needed to confirm this assertion. nist suppresses allosensitization in corneal transplantation. Arch Ophthalmol. 1998;116:1351–1357. Acknowledgments 21. Pape KA, Merica R, Mondino A, et al. Direct evidence that func- ϩ The authors thank IOVS volunteer editor Stephen Gee (University of tionally impaired CD4 T cells persist in vivo following induction of peripheral tolerance. J Immunol. 1998;160:4719–4729. North Carolina, Chapel Hill, NC) for editing the manuscript. 22. Schwartz RH. A cell culture model for T lymphocyte clonal anergy. References Science. 1990;248:1349–1356. 23. Chargui J, Hase T, Wada S, et al. NKT Cells as nonspecific immune- 1. The Collaborative Corneal Transplantation Studies Research regulator inducing tolerance in mouse model transplantation. Group. The collaborative corneal transplantation studies (CCTS). Transplant Proc. 2001;33:3833–3834.

Downloaded from iovs.arvojournals.org on 10/02/2021