- Email: [email protected]
Macrophage Migration Inhibition Factor Activity in the Aqueous Humor During Experimental Corneal Xenograft and Allograft Rejection
MACROPHAGE MIGRATION I N H I B I T I O N FACTOR ACTIVITY IN T H E AQUEOUS HUMOR DURING EXPERIMENTAL CORNEAL XENOGRAFT AND ALLOGRAFT REJECTION N E A L A. S H E R , M.D.,
D O N A L D J. D O U G H M A N ,
E L I Z A B E T H M I N D R U P , AND L L O Y D A. M I N A A I ,
M.D., M.D.
Minneapolis, Minnesota AND K E N N E T H A. F O O N ,
M.D.
Bethesda, Maryland
Immunologie rejection mediated by cellular immune components, remains an important cause of corneal graft failure. 1 In rabbits the resulting damage to the corneal graft, especially to the endothelium, is mediated by mononuclear cells, predominantly small lymphocytes. 2 The stimulus for the recruitment, migration, and activation of these and other effector cells into the graft is not understood but may possibly be regulated by soluble mediators of inflammation. These media tors include lymphokines, which are the product of the specific interaction be tween a sensitized lymphocyte and its sensitizing antigen. Large numbers of lymphokines have been identified. 3 One is the macrophage migration inhibition factor (MIF), characterized experimental ly 4 and in man. 5 M I F has correlated best with in vivo measurements of cellular hypersensitivity. 6 Sher and associates 7 · 8 reported on the feasibility of assaying for several media-
From the Department of Ophthalmology, Univer sity of Minnesota School of Medicine, Minneapolis, Minnesota, and the National Institutes of Health, Bethesda, Maryland. This study was supported in part by grants from the Minnesota Medical Founda tion, and National Institutes of Health grants EY0027 and EY-01211. Presented in part before the annual meeting of the Association for Research in Vision and Ophthalmo logy, April 26-30, 1976, Sarasota, Florida. Reprint requests to Neal A. Sher, M.D., Box 449 Mayo, University of Minnesota School of Medicine, Minneapolis, MN 55455.
tors of inflammation, including chemotactic factors and M I F , in the aqueous humor of rabbits during the course of immunogenic uveitis. In this study we demonstrate that significant M I F activity is measurable in the aqueous humor of rejecting xenografts and allografts. M A T E R I A L AND M E T H O D S
Adult female New Zealand (NZW) rab bits, weighing between 2 and 4 kg, were used as recipients in all experiments ex cept the interlamellar allografts in which 2- to 3-kg Dutch belted pigmented female rabbits were used. Anesthesia was ob tained with sodium pentobarbital and topical proparacaine hydrochloride (Ophthaine, 0.5%). All surgery was performed by one individual on one eye of each rabbit by using a Zeiss operating micro scope. Penetrating corneal xenografts were performed using a modification of the method of Khodadoust. 9 Five millimeter in diameter fresh chicken full-thickness corneal buttons were placed centrally with a running 10-0 Ethilon suture. The suture was left in place throughout this experiment. Control eyes were treated in an identi cal manner. Each control eye had a 5-mm in diameter corneal button trephined, re moved, rotated 180 degrees, and resutured into place. Except for 2% atropine, 10% phenylephrine solution, and polymyxin B sulfate ophthalmic ointment
858
VOL. 82, NO. 6
CORNEAL GRAFT REJECTION
(Polymyxin-Neomycin-Bacitracin Oint ment) used at surgery and on the first and second postoperative days, no other med ications were used. Eccentric interlamellar pocket xenografts were performed according to the method of Lorenzetti and Kaufman. 10 Fresh chicken or pig full-thickness 5-mm in diameter corneal buttons were inserted into the interlamellar pocket, 1 to 2 mm from the corneoscleral limbus, and the pocket was closed with two 8-0 silk su tures. Interlamellar allografts were performed in a similar manner. Dutch belted pigmented rabbits of both sexes were the recipients of a 5-mm in diameter NZW rabbit corneal button. Skin sensitization was performed on day 28 by the method of Warwick, 11 using a skin graft from the same donor. The sutures were removed when the blood vessels reached the graft. Control interlamellar allografts and xenografts were performed in an identical manner; however, no cornea was inserted. Aqueous humor was obtained by enter ing the corneoscleral limbus of the anes thetized rabbit with a No. 25 "scalp vein" needle (Butterfly) and 0.10 to 0.15 ml of fluid was removed. All samples of aqeuous humor were centrifuged at 400 g x 10 minutes and the supernatants were frozen to minus 20°C. Grading of the graft rejection—All ani mals were evaluated three times weekly by one observer in a single-masked fash ion using a slit lamp. Animals were ex cluded from the study if any operative or postoperative complications caused im mediate graft opacification, or if any sys temic illness may have impaired immunological competence. Less than 10% of the animals were excluded for these rea sons. To avoid confusion with nonimmunologic causes of graft failure, graft rejec tion was diagnosed only if the graft was clear postoperatively and at least seven
859
days had elapsed from the surgery to the onset of rejection. The grafts were evalu ated for clarity, vascularization, and iritis on a scale of 0 to 3, as modified from Baldwin and Borgmann 1 2 and Doughman and associates. 13 C L A R I T Y — 0 indicates clear, good iris and pupil detail; 1: slight haze, iris and pupil visible; 2: moderate haze, iris and pupil detail visible; and 3: opaque graft, no iris or pupil detail. VASCULARIZATION—0
indicates
no
vessels; 1: vessels to graft edge; 2: vessels into less than one half of graft; and 3: vessels into more than one half of graft. IRITIS—0 indicates no inflammation; 1: slight congestion of iris vessels (usually segmental); 2: mild to moderate conges tion of iris vessels and mild edema; and 3: marked congestion, dilation of iris ves sels, and moderate iris edema. Diagnosis of rejection—Rejection was diagnosed when a previously clear graft developed a score of 1 or more for clarity or iritis in conjunction with a score of 2 or more for vascularity. Rejection was also diagnosed in the presence of a classic rejection line or immune ring. Based on the clinical and histologie characteriza tion of rejection of the corneal grafts and controls, this scoring system readily dif ferentiated rejection from nonrejection. Production of other ocular inflammato ry states—Immunogenic uveitis was pro duced by the intravitreal injection of 0.1 ml of endotoxin-free human serum albu min into ten rabbit eyes. 7 Aqueous humor was obtained at the peak of the inflamma tion (days 12 to 14). Alkali burns were produced with I N NaOH applied for 30 seconds with a cotton-tipped applicator to the corneal surface. The intracorneal in jection of 0.03 ml of clove oil, according to the method of Liebowitz, Lass, and Kupferman, 14 produced severe corneal and moderate anterior chamber inflam mation. Corneal edema and minimal iritis were produced by the mechanical de-
860
AMERICAN JOURNAL OF OPHTHALMOLOGY
bridement of the corneal endothelium by using a small cyclodialysis spatula in serted through a 4-mm limbal incision and closed with two 8-0 silk sutures. Aqueous humor for these three models was obtained 24 hours after the inciting agent was applied. Multiple anterior chamber taps were performed on one eye on days 1, 5, 10, and 12. The aqueous humor for the MIF assay was obtained from the tap taken on day 12. Assay for MIF—Certain modifications of the standard MIF assay were em ployed.15,16 Guinea pig peritoneal exudate cells (60 x 106/ml) suspended in medium (RPMI 1640) were drawn up in 50-μ1 capillary tubes and one end was sealed with clay. They were spun into a pellet at 130 X g for five minutes. The tubes were cut at the cell-fluid interface and fixed to the bottom of 16-mm tissue culture wells with sterile silicone grease. Appropriate dilutions of the aqueous
DECEMBER, 1976
humor in this medium containing 10% heat-inactivated guinea pig serum was added to each well and incubated at 37°C in 5% C 0 2 for 24 hours. The areas of migration were projected 20 times enlarged, traced, and measured with a planimeter. The areas from quadriplicate tubes were averaged and the re sults were expressed as the percentage of migration inhibition, where: migration inhibition (%) = mean area of migration" with aqueous humor ~~ mean area ot migration x 100. with media
RESULTS
Penetrating xenograft—Six chicken to rabbit xenografts and six control proce dures were performed. The course of the rejection and the MIF activity in the aqueous humor were observed (Fig. 1). There was some postoperative iritis in
Fig. 1 (Sher and associates). Mi gration inhibition factor (MIF) in the aqueous humor during rejection of penetrating xenografts in the rab bit. The course of rejection in six penetrating chicken to rabbit grafts and six control procedures is shown in the upper and middle panels. The M I F activity, measured at a dilution of 1:10, is shown in the bottom panel.
VOL. 82, NO. 6
CORNEAL GRAFT REJECTION
both the experimental and control ani mals. Vascularization occurred in all the control grafts due to the presence of the sutures. We measured increased M I F activity at the first sign of graft rejection (day 9) and twice during the course of the rejection. No significant M I F activity was measura ble by day 65. We measured minimal M I F activity in the aqueous humor from the control eyes only at the earliest inter val and not during the peak of the rejec tion. Interlamellar xenograft—To deter mine whether M I F activity could be mea sured when the xenograft was not in di rect contact with the anterior chamber, interlamellar chicken and pig xenografts were performed and we measured the M I F activity. Five interlamellar and five control chicken grafts were monitored (Fig. 2). The first signs of rejection were observed on day 9. Aqueous humor was also ob tained on day 13 when the rejection was more severe. The M I F assay was per formed twice at a dilution of 1:10 (Fig. 2). M I F activity increased from 28% during early rejection to 9 1 % by day 13, parallel ing the increase in clinical severity of the rejection. Five interlamellar and five control pig grafts were monitored. Aqueous humor was obtained at day 10, early in the course of the rejection, revealing mod erate M I F activity. Interlamellar allograft—Ten allografts were performed and follow-up was ob tained in nine eyes (Fig. 3). Rejection of the skin grafts occurred on day 36 ± 2; rejection of the corneal allografts oc curred on day 47 ± 2. There was minimal iritis during the rejection. There was no iritis or clouding in any of the control grafts. To demonstrate the individual varia tion during the course of the corneal allograft rejection, the course and M I F activity from four individual rabbits are
861
Fig. 2 (Sher and associates). Migration inhibition factor (MIF) activity in the aqueous humor during rejection of chicken or pig into rabbit interlamellar pocket xenografts. The graft reactions for five chick en to rabbit grafts on days 9 and 13 and five pig to rabbit grafts (day 10) are shown in the upper panel. MIF activity measured at a dilution of 1:10 is shown for the controls and xenografts.
presented (Fig. 4). The aqueous humor samples were diluted with the medium (1:5) in this assay. The aqueous humor from five control rabbits (obtained on days 42, and 60) were pooled and used as controls; these results represent M I F levels of less than 10%. There was marked individual variation among the courses of the rejection and corresponding M I F activity (Fig. 4). Rab bit No. 53 had an early fulminant rejec tion with some mild iritis, and high levels of M I F were measured early in the rejec tion. Rabbit No. 49, with its milder rejec tion, did not have measurable M I F at this early stage. In all cases, M I F activity in the aqueous humor was present sometime during the rejection. These data demonstrated that M I F ac tivity is present in the aqueous humor during corneal graft rejection. To deter mine if the production of M I F activity is restricted to inflammations mediated by specific immunologie mechanisms, such
862
AMERICAN JOURNAL OF OPHTHALMOLOGY
DECEMBER, 1976
Days after Transplantation Fig. 3 (Sher and associates). Summary of course of rejection of interlamellar allografts. The course of the graft reaction for nine allografts is summarized. The mean ± S.E. and the range (shaded) of the time of skin and corneal graft rejection are indicated.
as xenograft and allograft rejections and immunogenic uveitis, or is present in any nonspecific ocular inflammation, several nonspecific inflammations were studied: multiple paracentesis procedures, corneal alkali burns, clove oil-induced corneal inflammation, and mechanical debridement of the corneal endothelium (Table). Groups of two or three rabbits were used for each model except immunogenic uvei tis, in which ten animals were used. The aqueous humor was obtained and treated in an identical manner as in the corneal graft experiments. No significant M I F activity was measured in any of the nonimmunologic inflammations. Increased titers of MIF activity were present in immunogenic uveitis. DISCUSSION
Migration inhibition factor activity is present in the aqueous humor of both rejecting corneal xenografts and allo grafts. MIF activity was measured at the
earliest visible signs of clinical rejection, correlated at times with the severity of the graft reaction, and was not measurable in the absence of inflammation mediated by immunologie mechanisms. We did not obtain aqueous humor to test for M I F activity before clinical evidence of rejec tion because the paracentesis procedure itself may cloud the graft secondary to endothelial trauma. In addition, the para centesis procedure may induce nonspe cific inflammation via the release of various mediators of inflammation in cluding prostaglandins. 1 7 Nonspecific inflammation can be a cause of graft failure. 18 We used the term M I F activity to refer to the partially characterized glycoprotein lymphocyte product. 4 , 5 The activity we have described is probably due to MIF. However, to characterize this factor, large volumes of aqueous humor would have to be collected. To our knowledge, this is the first re-
863
CORNEAL GRAFT REJECTION
VOL. 82, NO. 6
Clinical Score
MIF Activity of Aqueous Humor
% Inhibition of Macrophage Migration -//t
t REJ. SKIN-"' TRANS PLANTATION
3.0
• ··. / Vasculor• f& nation ■ / \ Clarity
1.5
0.5
Ï
Ll f CORNEAL GRAFT REJECTION
*
y
P
RABBIT054
/S y
/ Vaxukriiation
h
/ v
—o — ^ A o o o d
b <- Iritis
! ί \ Clarity
—
y
too^>o
Il V .·
o
/
— er
0
p
^-Iritis
MIF Activity of Aqueous Humor
80
% Inhibition of Macrophage Migration
I
Clinical Grading ~ RABBITS53
2.0
1.0
I
tREJ. SKIN-" TRANSPLANTATION
CORNEAL GRAFT REJECTION
2.5
Clinical Score
l
4 5 6 7 8 9
60 40 20 0
J
//I Ill I I ■! 4 5 6 7 8 9 tREJ. f
SKIN TRANSPLANTATION
CORNEAL GRAFT REJECTION
i
i
-il· 4 5
tREJ. SKIN TRANS PLANTATION
i
i
7
t
1
8 9
CORNEAL GRAFT REJECTION
Weeks after Corneal Transplantation Fig. 4 (Sher and associates). Migration inhibition factor (MIF) activity of the aqueous humor during rejection of the interlamellar pocket allografts. The course of the rejection and the M I F activity measured at a dilution of 1:5 for four rabbits (Nos. 49, 50, 53, and 54) are shown. The results of the pooled control aqueous humor samples, taken at days 48 and 60, are illustrated with rabbit No. 49.
864
AMERICAN JOURNAL OF OPHTHALMOLOGY TABLE
M I F ACTIVITY IN OTHER OCULAR INFLAMMATORY CONDITIONS
Type of Inflammation Multiple paracentesis procedures Removal of corneal endothelium Alkali burn Intracorneal clove oil Immunogenic uveitis
Dilution of Aqueous Humor
MIF Activity, %
1:5
0
1:5
18
1:5 1:5 1:8 1:16
0* Of 74 43
*Actual value for MIF was 127% of control migra tion, represented as 0. t Actual value for MIF was 130% of control migra tion, represented as 0.
port of the measurement of M I F activity in the aqueous humor during graft rejec tion. Its presence is predictable since sev eral studies have shown that in various pathologic states, the aqueous humor contains elevated levels of complement components, 1 9 prostaglandins, 2 0 chemotactic factors, 7,8 and factors that stimulate lymphocyte transformation. 21 The site of M I F production and how it gains access to the aqueous humor is not known. Local and systemic proliferation of lymphoid tissue after a corneal xenograft, 22 as well as sensitization of lymphocytes to corneal antigens, 23,24 have been demonstrated. Whether the effector lymphocytes are derived from the uveal tract and gain access to the graft via the anterior cham ber or from limbal vessels is unknown. Our data suggest that the mediators re sponsible for the M I F activity, originat ing either from the graft bed, corneoscleral limbus, or uveal tract, eventually diffuse into the aqueous humor. It would not be surprising to find M I F activity in the vitreous or tears, or both.
DECEMBER, 1976
The role of antibodies in mediating graft rejection is generally thought to be secondary to cellular immunity. Howev er, both the experimental xenograft and allograft may induce antibody forma tion. 25 In special circumstances, antigenantibody complexes have inhibited mac rophage migration. 26 The possibility that the M I F activity may be mediated in part by this mechanism cannot be ruled out. A clinical problem in corneal trans plantation is the lack of a specific diag nostic test to aid in distinguishing a true immunologie rejection from graft failure due to other causes, such as poor viability of donor material or recurrence of the underlying disease. The end result, in some cases, is an opacified vascularized graft. The absence of significant M I F ac tivity in various types of experimental nonspecific inflammations implicates an immunologie mechanism in the elabora tion of MIF. These data suggest that the assay for M I F in the aqueous humor may provide additional diagnostic informa tion as to the presence of a true immuno logie rejection of the corneal graft. The study of various immunologie mediators in the accessible ocular fluids may pro vide a better understanding of the pathophysiology of ocular inflammation and graft rejection. SUMMARY
Macrophage migration inhibition fac tor (MIF), a soluble mediator of delayed hypersensitivity, was assayed for in the aqueous humor of rabbits undergoing corneal graft rejection. Penetrating and interlamellar xenografts and interlamellar allografts were performed in rabbits, and aqueous humor was obtained during the rejection. Significant levels of M I F activi ty were measured in the aqueous humor early in the course of xenograft and allo graft rejection, and M I F activity was present during the course of the active
VOL. 82, NO. 6
CORNEAL GRAFT REJECTION
rejection. This activity returned to near normal after the active rejection re solved. No significant MIF activity could be measured during the nonspecific inflammations produced by alkali burns, multiple paracentèses, intracorneal clove oil, or mechanical debriding of the endothelium. REFERENCES 1. Polack, F. M.: Clinical and pathologic aspects of the comeal graft reaction. Trans. Am. Acad. Ophthalmol. Otolaryngol. 77:418, 1973. 2. Khodadoust, A. A., and Silverstein, A. M.: Induction of corneal graft rejection by passive cell transfer. Invest. Ophthalmol. 15:89, 1976. 3. David, J. R., and David, R. R.: Cellular hyper sensitivity and immunity. Inhibition of macrophage migration and the lymphocyte mediators. Prog. Al lergy 16:300, 1972. 4. Bennett, B., and Bloom, B. R.: Studies on the migration inhibitory factors associated with delayed-type hypersensitivity. Cytodynamics and specificity. Transplantation 5:996, 1967. 5. Rocklin, R. E., Remold, H. G., and David, J. R.: Characterization of human migration inhibitory fac tor (MIF) from antigen-stimulated lymphocytes. Cell. Immunol. 5:436, 1972. 6. Bloom, B. R., and Bennett, B.: Relation of the migration inhibitory factor (MIF) to delayed type hypersensitivity reactions. Ann. N. Y. Acad. Sei. 169:258, 1970. 7. Sher, N. A., Foon, K. A., Fishman, M., and Brown, T. M.: Demonstration of macrophage chemotactic factors in the aqueous humor during experimental immunogenic uveitis. Infect. Immun. 13:1110, 1976. 8. Foon, K. A., Sher, N. A., and Fishman, M. L.: Macrophage chemotactic factor and migration inhi bition factor in the aqueous humor during experi mental uveitis. Fed. Proc. 35:492, 1976. 9. Khodadoust, A. A.: Penetrating keratoplasty in the rabbit. Am. J. Ophthalmol. 66:899, 1968. 10. Lorenzetti, D. W. C , and Kaufman, H. E.: Experimental production of graft reactions with suppression by topical corticosteroids. Arch. Oph thalmol. 76:274, 1966. 11. Warwick, W. J.: A rapid sutureless method for grafting skin in rabbits. Transplantation Bull. 30:51, 1962. 12. Baldwin, H. A., and Borgmann, A. P.: An improved immunogenic uveitis test in rabbits for
865
evaluation of anti-inflammatory compounds. Proc. Soc. Exp. Biol. 133:1326,1970. 13. Doughman, D. J., Miller, G. E., Mindrup, E. A., Schmitt, M. K., Harris, J. E., and Good, R. A.: The fate of experimental organ cultured corneal xenografts. Transplantation, in press. 14.Liebowitz, H. M., Lass, J. H., and Kupferman, A.: Quantitation of inflammation in the cornea. Arch. Ophthalmol. 92:427, 1974. 15. David, J. R., Al-Askari, S., Lawrence, H. S., and Thomas, L.: The specificity of inhibition of cell migrations by antigens. J. Immunol. 93:264, 1964. 16. Rosenstreich, D. L., Blake, T., and Rosenthal, A. S.: The peritoneal exudate lymphocyte. 1. Differ ences in antigen responsiveness between peritoneal exudate and lymph node lymphocytes from immu nized guinea pigs. J. Exp. Med. 134:1170, 1971. 17. Miller, J. D., Eakins, K. E., and Atwal, M.: The release of PGE2-like activity into aqueous humor after paracentesis and its prevention by aspi rin. Invest. Ophthalmol. 12:939, 1973. 18. Polack, F. M.: The effect of ocular inflamma tion on corneal grafts. Am. J. Ophthalmol. 60:259, 1965. 19. Chandler, J. W., Leder, R., Kaufman, H. E., and Caldwell, J. R.: Quantitative determination of complement components and immunoglobulins in tears and aqueous humor. Invest. Ophthalmol. 13: 151, 1974. 20. Eakins, K. E., Whitelock, R. A., Perkins, E. S., Bennett, A., and Unger, W. G.: Release of prostaglandins in ocular inflammation in the rabbit. Na ture (New Biol.) 239:248, 1972. 21. Beneyra, D., and Sachs, V.: Growth factors in aqueous humor of normal and inflamed eyes of rabbits. Invest. Ophthalmol. 13:868, 1974. 22. Polack, F. M., and Gonzales, C. E.: The response of the lymphoid tissue to corneal heterografts. Arch. Ophthalmol. 80:321, 1968. 23. Wang, H. S., and Basu, P. K.: Cellular immu nity to xenogeneic corneal grafts in rabbits. Can. J. Ophthalmol. 10:263, 1975. 24. Basu, P. K., and Carré, F.: Involvement of immune lymphocytes in corneal reaction against foreign tissue antigen as determined by leukocyte migration inhibition test. Can J. Ophthalmol. 9:236, 1974. 25. D'Ermo, F., Lanzieri, M., and Secchi, A.: Anticomeal antibodies in rabbits after homologous and heterologous comeal grafts. Acta Ophthalmol. 44:233, 1966. 26. Kotkes, P., and Pick, E.: Studies on the inhibi tion of macrophage migration induced by soluble antigen-antibody complexes. Clin. Exp. Immunol. 19:105, 1975.
D O N A L D J. D O U G H M A N ,
E L I Z A B E T H M I N D R U P , AND L L O Y D A. M I N A A I ,
M.D., M.D.
Minneapolis, Minnesota AND K E N N E T H A. F O O N ,
M.D.
Bethesda, Maryland
Immunologie rejection mediated by cellular immune components, remains an important cause of corneal graft failure. 1 In rabbits the resulting damage to the corneal graft, especially to the endothelium, is mediated by mononuclear cells, predominantly small lymphocytes. 2 The stimulus for the recruitment, migration, and activation of these and other effector cells into the graft is not understood but may possibly be regulated by soluble mediators of inflammation. These media tors include lymphokines, which are the product of the specific interaction be tween a sensitized lymphocyte and its sensitizing antigen. Large numbers of lymphokines have been identified. 3 One is the macrophage migration inhibition factor (MIF), characterized experimental ly 4 and in man. 5 M I F has correlated best with in vivo measurements of cellular hypersensitivity. 6 Sher and associates 7 · 8 reported on the feasibility of assaying for several media-
From the Department of Ophthalmology, Univer sity of Minnesota School of Medicine, Minneapolis, Minnesota, and the National Institutes of Health, Bethesda, Maryland. This study was supported in part by grants from the Minnesota Medical Founda tion, and National Institutes of Health grants EY0027 and EY-01211. Presented in part before the annual meeting of the Association for Research in Vision and Ophthalmo logy, April 26-30, 1976, Sarasota, Florida. Reprint requests to Neal A. Sher, M.D., Box 449 Mayo, University of Minnesota School of Medicine, Minneapolis, MN 55455.
tors of inflammation, including chemotactic factors and M I F , in the aqueous humor of rabbits during the course of immunogenic uveitis. In this study we demonstrate that significant M I F activity is measurable in the aqueous humor of rejecting xenografts and allografts. M A T E R I A L AND M E T H O D S
Adult female New Zealand (NZW) rab bits, weighing between 2 and 4 kg, were used as recipients in all experiments ex cept the interlamellar allografts in which 2- to 3-kg Dutch belted pigmented female rabbits were used. Anesthesia was ob tained with sodium pentobarbital and topical proparacaine hydrochloride (Ophthaine, 0.5%). All surgery was performed by one individual on one eye of each rabbit by using a Zeiss operating micro scope. Penetrating corneal xenografts were performed using a modification of the method of Khodadoust. 9 Five millimeter in diameter fresh chicken full-thickness corneal buttons were placed centrally with a running 10-0 Ethilon suture. The suture was left in place throughout this experiment. Control eyes were treated in an identi cal manner. Each control eye had a 5-mm in diameter corneal button trephined, re moved, rotated 180 degrees, and resutured into place. Except for 2% atropine, 10% phenylephrine solution, and polymyxin B sulfate ophthalmic ointment
858
VOL. 82, NO. 6
CORNEAL GRAFT REJECTION
(Polymyxin-Neomycin-Bacitracin Oint ment) used at surgery and on the first and second postoperative days, no other med ications were used. Eccentric interlamellar pocket xenografts were performed according to the method of Lorenzetti and Kaufman. 10 Fresh chicken or pig full-thickness 5-mm in diameter corneal buttons were inserted into the interlamellar pocket, 1 to 2 mm from the corneoscleral limbus, and the pocket was closed with two 8-0 silk su tures. Interlamellar allografts were performed in a similar manner. Dutch belted pigmented rabbits of both sexes were the recipients of a 5-mm in diameter NZW rabbit corneal button. Skin sensitization was performed on day 28 by the method of Warwick, 11 using a skin graft from the same donor. The sutures were removed when the blood vessels reached the graft. Control interlamellar allografts and xenografts were performed in an identical manner; however, no cornea was inserted. Aqueous humor was obtained by enter ing the corneoscleral limbus of the anes thetized rabbit with a No. 25 "scalp vein" needle (Butterfly) and 0.10 to 0.15 ml of fluid was removed. All samples of aqeuous humor were centrifuged at 400 g x 10 minutes and the supernatants were frozen to minus 20°C. Grading of the graft rejection—All ani mals were evaluated three times weekly by one observer in a single-masked fash ion using a slit lamp. Animals were ex cluded from the study if any operative or postoperative complications caused im mediate graft opacification, or if any sys temic illness may have impaired immunological competence. Less than 10% of the animals were excluded for these rea sons. To avoid confusion with nonimmunologic causes of graft failure, graft rejec tion was diagnosed only if the graft was clear postoperatively and at least seven
859
days had elapsed from the surgery to the onset of rejection. The grafts were evalu ated for clarity, vascularization, and iritis on a scale of 0 to 3, as modified from Baldwin and Borgmann 1 2 and Doughman and associates. 13 C L A R I T Y — 0 indicates clear, good iris and pupil detail; 1: slight haze, iris and pupil visible; 2: moderate haze, iris and pupil detail visible; and 3: opaque graft, no iris or pupil detail. VASCULARIZATION—0
indicates
no
vessels; 1: vessels to graft edge; 2: vessels into less than one half of graft; and 3: vessels into more than one half of graft. IRITIS—0 indicates no inflammation; 1: slight congestion of iris vessels (usually segmental); 2: mild to moderate conges tion of iris vessels and mild edema; and 3: marked congestion, dilation of iris ves sels, and moderate iris edema. Diagnosis of rejection—Rejection was diagnosed when a previously clear graft developed a score of 1 or more for clarity or iritis in conjunction with a score of 2 or more for vascularity. Rejection was also diagnosed in the presence of a classic rejection line or immune ring. Based on the clinical and histologie characteriza tion of rejection of the corneal grafts and controls, this scoring system readily dif ferentiated rejection from nonrejection. Production of other ocular inflammato ry states—Immunogenic uveitis was pro duced by the intravitreal injection of 0.1 ml of endotoxin-free human serum albu min into ten rabbit eyes. 7 Aqueous humor was obtained at the peak of the inflamma tion (days 12 to 14). Alkali burns were produced with I N NaOH applied for 30 seconds with a cotton-tipped applicator to the corneal surface. The intracorneal in jection of 0.03 ml of clove oil, according to the method of Liebowitz, Lass, and Kupferman, 14 produced severe corneal and moderate anterior chamber inflam mation. Corneal edema and minimal iritis were produced by the mechanical de-
860
AMERICAN JOURNAL OF OPHTHALMOLOGY
bridement of the corneal endothelium by using a small cyclodialysis spatula in serted through a 4-mm limbal incision and closed with two 8-0 silk sutures. Aqueous humor for these three models was obtained 24 hours after the inciting agent was applied. Multiple anterior chamber taps were performed on one eye on days 1, 5, 10, and 12. The aqueous humor for the MIF assay was obtained from the tap taken on day 12. Assay for MIF—Certain modifications of the standard MIF assay were em ployed.15,16 Guinea pig peritoneal exudate cells (60 x 106/ml) suspended in medium (RPMI 1640) were drawn up in 50-μ1 capillary tubes and one end was sealed with clay. They were spun into a pellet at 130 X g for five minutes. The tubes were cut at the cell-fluid interface and fixed to the bottom of 16-mm tissue culture wells with sterile silicone grease. Appropriate dilutions of the aqueous
DECEMBER, 1976
humor in this medium containing 10% heat-inactivated guinea pig serum was added to each well and incubated at 37°C in 5% C 0 2 for 24 hours. The areas of migration were projected 20 times enlarged, traced, and measured with a planimeter. The areas from quadriplicate tubes were averaged and the re sults were expressed as the percentage of migration inhibition, where: migration inhibition (%) = mean area of migration" with aqueous humor ~~ mean area ot migration x 100. with media
RESULTS
Penetrating xenograft—Six chicken to rabbit xenografts and six control proce dures were performed. The course of the rejection and the MIF activity in the aqueous humor were observed (Fig. 1). There was some postoperative iritis in
Fig. 1 (Sher and associates). Mi gration inhibition factor (MIF) in the aqueous humor during rejection of penetrating xenografts in the rab bit. The course of rejection in six penetrating chicken to rabbit grafts and six control procedures is shown in the upper and middle panels. The M I F activity, measured at a dilution of 1:10, is shown in the bottom panel.
VOL. 82, NO. 6
CORNEAL GRAFT REJECTION
both the experimental and control ani mals. Vascularization occurred in all the control grafts due to the presence of the sutures. We measured increased M I F activity at the first sign of graft rejection (day 9) and twice during the course of the rejection. No significant M I F activity was measura ble by day 65. We measured minimal M I F activity in the aqueous humor from the control eyes only at the earliest inter val and not during the peak of the rejec tion. Interlamellar xenograft—To deter mine whether M I F activity could be mea sured when the xenograft was not in di rect contact with the anterior chamber, interlamellar chicken and pig xenografts were performed and we measured the M I F activity. Five interlamellar and five control chicken grafts were monitored (Fig. 2). The first signs of rejection were observed on day 9. Aqueous humor was also ob tained on day 13 when the rejection was more severe. The M I F assay was per formed twice at a dilution of 1:10 (Fig. 2). M I F activity increased from 28% during early rejection to 9 1 % by day 13, parallel ing the increase in clinical severity of the rejection. Five interlamellar and five control pig grafts were monitored. Aqueous humor was obtained at day 10, early in the course of the rejection, revealing mod erate M I F activity. Interlamellar allograft—Ten allografts were performed and follow-up was ob tained in nine eyes (Fig. 3). Rejection of the skin grafts occurred on day 36 ± 2; rejection of the corneal allografts oc curred on day 47 ± 2. There was minimal iritis during the rejection. There was no iritis or clouding in any of the control grafts. To demonstrate the individual varia tion during the course of the corneal allograft rejection, the course and M I F activity from four individual rabbits are
861
Fig. 2 (Sher and associates). Migration inhibition factor (MIF) activity in the aqueous humor during rejection of chicken or pig into rabbit interlamellar pocket xenografts. The graft reactions for five chick en to rabbit grafts on days 9 and 13 and five pig to rabbit grafts (day 10) are shown in the upper panel. MIF activity measured at a dilution of 1:10 is shown for the controls and xenografts.
presented (Fig. 4). The aqueous humor samples were diluted with the medium (1:5) in this assay. The aqueous humor from five control rabbits (obtained on days 42, and 60) were pooled and used as controls; these results represent M I F levels of less than 10%. There was marked individual variation among the courses of the rejection and corresponding M I F activity (Fig. 4). Rab bit No. 53 had an early fulminant rejec tion with some mild iritis, and high levels of M I F were measured early in the rejec tion. Rabbit No. 49, with its milder rejec tion, did not have measurable M I F at this early stage. In all cases, M I F activity in the aqueous humor was present sometime during the rejection. These data demonstrated that M I F ac tivity is present in the aqueous humor during corneal graft rejection. To deter mine if the production of M I F activity is restricted to inflammations mediated by specific immunologie mechanisms, such
862
AMERICAN JOURNAL OF OPHTHALMOLOGY
DECEMBER, 1976
Days after Transplantation Fig. 3 (Sher and associates). Summary of course of rejection of interlamellar allografts. The course of the graft reaction for nine allografts is summarized. The mean ± S.E. and the range (shaded) of the time of skin and corneal graft rejection are indicated.
as xenograft and allograft rejections and immunogenic uveitis, or is present in any nonspecific ocular inflammation, several nonspecific inflammations were studied: multiple paracentesis procedures, corneal alkali burns, clove oil-induced corneal inflammation, and mechanical debridement of the corneal endothelium (Table). Groups of two or three rabbits were used for each model except immunogenic uvei tis, in which ten animals were used. The aqueous humor was obtained and treated in an identical manner as in the corneal graft experiments. No significant M I F activity was measured in any of the nonimmunologic inflammations. Increased titers of MIF activity were present in immunogenic uveitis. DISCUSSION
Migration inhibition factor activity is present in the aqueous humor of both rejecting corneal xenografts and allo grafts. MIF activity was measured at the
earliest visible signs of clinical rejection, correlated at times with the severity of the graft reaction, and was not measurable in the absence of inflammation mediated by immunologie mechanisms. We did not obtain aqueous humor to test for M I F activity before clinical evidence of rejec tion because the paracentesis procedure itself may cloud the graft secondary to endothelial trauma. In addition, the para centesis procedure may induce nonspe cific inflammation via the release of various mediators of inflammation in cluding prostaglandins. 1 7 Nonspecific inflammation can be a cause of graft failure. 18 We used the term M I F activity to refer to the partially characterized glycoprotein lymphocyte product. 4 , 5 The activity we have described is probably due to MIF. However, to characterize this factor, large volumes of aqueous humor would have to be collected. To our knowledge, this is the first re-
863
CORNEAL GRAFT REJECTION
VOL. 82, NO. 6
Clinical Score
MIF Activity of Aqueous Humor
% Inhibition of Macrophage Migration -//t
t REJ. SKIN-"' TRANS PLANTATION
3.0
• ··. / Vasculor• f& nation ■ / \ Clarity
1.5
0.5
Ï
Ll f CORNEAL GRAFT REJECTION
*
y
P
RABBIT054
/S y
/ Vaxukriiation
h
/ v
—o — ^ A o o o d
b <- Iritis
! ί \ Clarity
—
y
too^>o
Il V .·
o
/
— er
0
p
^-Iritis
MIF Activity of Aqueous Humor
80
% Inhibition of Macrophage Migration
I
Clinical Grading ~ RABBITS53
2.0
1.0
I
tREJ. SKIN-" TRANSPLANTATION
CORNEAL GRAFT REJECTION
2.5
Clinical Score
l
4 5 6 7 8 9
60 40 20 0
J
//I Ill I I ■! 4 5 6 7 8 9 tREJ. f
SKIN TRANSPLANTATION
CORNEAL GRAFT REJECTION
i
i
-il· 4 5
tREJ. SKIN TRANS PLANTATION
i
i
7
t
1
8 9
CORNEAL GRAFT REJECTION
Weeks after Corneal Transplantation Fig. 4 (Sher and associates). Migration inhibition factor (MIF) activity of the aqueous humor during rejection of the interlamellar pocket allografts. The course of the rejection and the M I F activity measured at a dilution of 1:5 for four rabbits (Nos. 49, 50, 53, and 54) are shown. The results of the pooled control aqueous humor samples, taken at days 48 and 60, are illustrated with rabbit No. 49.
864
AMERICAN JOURNAL OF OPHTHALMOLOGY TABLE
M I F ACTIVITY IN OTHER OCULAR INFLAMMATORY CONDITIONS
Type of Inflammation Multiple paracentesis procedures Removal of corneal endothelium Alkali burn Intracorneal clove oil Immunogenic uveitis
Dilution of Aqueous Humor
MIF Activity, %
1:5
0
1:5
18
1:5 1:5 1:8 1:16
0* Of 74 43
*Actual value for MIF was 127% of control migra tion, represented as 0. t Actual value for MIF was 130% of control migra tion, represented as 0.
port of the measurement of M I F activity in the aqueous humor during graft rejec tion. Its presence is predictable since sev eral studies have shown that in various pathologic states, the aqueous humor contains elevated levels of complement components, 1 9 prostaglandins, 2 0 chemotactic factors, 7,8 and factors that stimulate lymphocyte transformation. 21 The site of M I F production and how it gains access to the aqueous humor is not known. Local and systemic proliferation of lymphoid tissue after a corneal xenograft, 22 as well as sensitization of lymphocytes to corneal antigens, 23,24 have been demonstrated. Whether the effector lymphocytes are derived from the uveal tract and gain access to the graft via the anterior cham ber or from limbal vessels is unknown. Our data suggest that the mediators re sponsible for the M I F activity, originat ing either from the graft bed, corneoscleral limbus, or uveal tract, eventually diffuse into the aqueous humor. It would not be surprising to find M I F activity in the vitreous or tears, or both.
DECEMBER, 1976
The role of antibodies in mediating graft rejection is generally thought to be secondary to cellular immunity. Howev er, both the experimental xenograft and allograft may induce antibody forma tion. 25 In special circumstances, antigenantibody complexes have inhibited mac rophage migration. 26 The possibility that the M I F activity may be mediated in part by this mechanism cannot be ruled out. A clinical problem in corneal trans plantation is the lack of a specific diag nostic test to aid in distinguishing a true immunologie rejection from graft failure due to other causes, such as poor viability of donor material or recurrence of the underlying disease. The end result, in some cases, is an opacified vascularized graft. The absence of significant M I F ac tivity in various types of experimental nonspecific inflammations implicates an immunologie mechanism in the elabora tion of MIF. These data suggest that the assay for M I F in the aqueous humor may provide additional diagnostic informa tion as to the presence of a true immuno logie rejection of the corneal graft. The study of various immunologie mediators in the accessible ocular fluids may pro vide a better understanding of the pathophysiology of ocular inflammation and graft rejection. SUMMARY
Macrophage migration inhibition fac tor (MIF), a soluble mediator of delayed hypersensitivity, was assayed for in the aqueous humor of rabbits undergoing corneal graft rejection. Penetrating and interlamellar xenografts and interlamellar allografts were performed in rabbits, and aqueous humor was obtained during the rejection. Significant levels of M I F activi ty were measured in the aqueous humor early in the course of xenograft and allo graft rejection, and M I F activity was present during the course of the active
VOL. 82, NO. 6
CORNEAL GRAFT REJECTION
rejection. This activity returned to near normal after the active rejection re solved. No significant MIF activity could be measured during the nonspecific inflammations produced by alkali burns, multiple paracentèses, intracorneal clove oil, or mechanical debriding of the endothelium. REFERENCES 1. Polack, F. M.: Clinical and pathologic aspects of the comeal graft reaction. Trans. Am. Acad. Ophthalmol. Otolaryngol. 77:418, 1973. 2. Khodadoust, A. A., and Silverstein, A. M.: Induction of corneal graft rejection by passive cell transfer. Invest. Ophthalmol. 15:89, 1976. 3. David, J. R., and David, R. R.: Cellular hyper sensitivity and immunity. Inhibition of macrophage migration and the lymphocyte mediators. Prog. Al lergy 16:300, 1972. 4. Bennett, B., and Bloom, B. R.: Studies on the migration inhibitory factors associated with delayed-type hypersensitivity. Cytodynamics and specificity. Transplantation 5:996, 1967. 5. Rocklin, R. E., Remold, H. G., and David, J. R.: Characterization of human migration inhibitory fac tor (MIF) from antigen-stimulated lymphocytes. Cell. Immunol. 5:436, 1972. 6. Bloom, B. R., and Bennett, B.: Relation of the migration inhibitory factor (MIF) to delayed type hypersensitivity reactions. Ann. N. Y. Acad. Sei. 169:258, 1970. 7. Sher, N. A., Foon, K. A., Fishman, M., and Brown, T. M.: Demonstration of macrophage chemotactic factors in the aqueous humor during experimental immunogenic uveitis. Infect. Immun. 13:1110, 1976. 8. Foon, K. A., Sher, N. A., and Fishman, M. L.: Macrophage chemotactic factor and migration inhi bition factor in the aqueous humor during experi mental uveitis. Fed. Proc. 35:492, 1976. 9. Khodadoust, A. A.: Penetrating keratoplasty in the rabbit. Am. J. Ophthalmol. 66:899, 1968. 10. Lorenzetti, D. W. C , and Kaufman, H. E.: Experimental production of graft reactions with suppression by topical corticosteroids. Arch. Oph thalmol. 76:274, 1966. 11. Warwick, W. J.: A rapid sutureless method for grafting skin in rabbits. Transplantation Bull. 30:51, 1962. 12. Baldwin, H. A., and Borgmann, A. P.: An improved immunogenic uveitis test in rabbits for
865
evaluation of anti-inflammatory compounds. Proc. Soc. Exp. Biol. 133:1326,1970. 13. Doughman, D. J., Miller, G. E., Mindrup, E. A., Schmitt, M. K., Harris, J. E., and Good, R. A.: The fate of experimental organ cultured corneal xenografts. Transplantation, in press. 14.Liebowitz, H. M., Lass, J. H., and Kupferman, A.: Quantitation of inflammation in the cornea. Arch. Ophthalmol. 92:427, 1974. 15. David, J. R., Al-Askari, S., Lawrence, H. S., and Thomas, L.: The specificity of inhibition of cell migrations by antigens. J. Immunol. 93:264, 1964. 16. Rosenstreich, D. L., Blake, T., and Rosenthal, A. S.: The peritoneal exudate lymphocyte. 1. Differ ences in antigen responsiveness between peritoneal exudate and lymph node lymphocytes from immu nized guinea pigs. J. Exp. Med. 134:1170, 1971. 17. Miller, J. D., Eakins, K. E., and Atwal, M.: The release of PGE2-like activity into aqueous humor after paracentesis and its prevention by aspi rin. Invest. Ophthalmol. 12:939, 1973. 18. Polack, F. M.: The effect of ocular inflamma tion on corneal grafts. Am. J. Ophthalmol. 60:259, 1965. 19. Chandler, J. W., Leder, R., Kaufman, H. E., and Caldwell, J. R.: Quantitative determination of complement components and immunoglobulins in tears and aqueous humor. Invest. Ophthalmol. 13: 151, 1974. 20. Eakins, K. E., Whitelock, R. A., Perkins, E. S., Bennett, A., and Unger, W. G.: Release of prostaglandins in ocular inflammation in the rabbit. Na ture (New Biol.) 239:248, 1972. 21. Beneyra, D., and Sachs, V.: Growth factors in aqueous humor of normal and inflamed eyes of rabbits. Invest. Ophthalmol. 13:868, 1974. 22. Polack, F. M., and Gonzales, C. E.: The response of the lymphoid tissue to corneal heterografts. Arch. Ophthalmol. 80:321, 1968. 23. Wang, H. S., and Basu, P. K.: Cellular immu nity to xenogeneic corneal grafts in rabbits. Can. J. Ophthalmol. 10:263, 1975. 24. Basu, P. K., and Carré, F.: Involvement of immune lymphocytes in corneal reaction against foreign tissue antigen as determined by leukocyte migration inhibition test. Can J. Ophthalmol. 9:236, 1974. 25. D'Ermo, F., Lanzieri, M., and Secchi, A.: Anticomeal antibodies in rabbits after homologous and heterologous comeal grafts. Acta Ophthalmol. 44:233, 1966. 26. Kotkes, P., and Pick, E.: Studies on the inhibi tion of macrophage migration induced by soluble antigen-antibody complexes. Clin. Exp. Immunol. 19:105, 1975.