Effect of local corticosteroids on antibody-forming cells in the eye and draining lymph nodes. R. F. Meyer, G. Smolin, J. M. Hall, and M.

Transcription

1 Effect of local corticosteroids on antibody-forming cells in the eye and draining lymph nodes R. F. Meyer, G. Smolin, J. M. Hall, and M. Okumoto Significant numbers of antibody-forming cells (AFC) have been found in the cornea, uvcal tract, and. draining lymph nodes after the intracorneal injection of bovine gamma-globulin (BGC). To study the effect of locally administered corticostcroids on these antibody-forming tissues, we made unilateral intracorncal injections of rabbit eyes with BGG. These we followed immediately with suhconjunctival injections of 1 mg. of triamcinolonc suspension, and then with a second round of 1 mg. injections seven days later. A control group of animals received the BGG injections followed by two subconjunctioal saline injections. We killed the animals on postinjection days 6,,,, and, and tested the draining lymph nodes, homolateral uvcal tissue, and homolateral cornea for AFC by a modification of the Jcrne placque technique. The local steroids had no effect on the number of AFC produced in the draining lymph nodes or on the circulating antibody response, but they reduced the number of AFC in the homolatcral iiveal tracts and corneas. Clinically there was less inflammatory response in the steroid-treated eyes than in the control eyes. The possible mechanisms by which corticosteroids achieve their anti-immunologic and anti-inflammatory benefits are discussed. I t has recently been shown that the early immunologic reaction to the intracorneal injection of rabbit eyes with bovine gamma-globulin (BGG) is a local response confined essentially to the draining lymph From the Francis I. Proctor Foundation for Research in Ophthalmology, and the Department of Ophthalmology, University of California, San Francisco, Calif. Supported in part by Grants EY-11 and EY from the National Institutes of Health, Bethesda, Md. Dr. Meyer was the recipient ot a fellowship from the Heed Foundation. Submitted for publication June 26, 17. Reprint requests: Dr. Gilbert Smolin, 3-S, University of California Medical Center, San Francisco, Calif nodes and ocular tissues. 1 Antibody-forming cells (AFC) have been detected in the homolateral lymph nodes, cornea, and uveal tract by means of the Jerne placque technique 1 ; and migration inhibitory factor (MIF), a cell-free factor produced by sensitized lymphocytes, has been released by cells harvested from the draining homolateral lymph nodes.'-' It would seem, therefore, that both humoral and cellular immune responses participate in the reaction. Corticosteroids are known to have specific effects on lymphoid tissues and are used widely in the treatment of a number of diseases caused by immunologic processes. Although the way the corticosteroids achieve their immunosuppressive effect is not known, it is possible that they Downloaded from iovs.arvojournals.org on 1/26/

2 Volume N Number 2 Effect of local corticosteroids 13 act by (1) exerting a local anti-inhammatory effect/ 1 (2) stabilizing the lysosomal membranes, 1 (3) destroying the lymphocytes,' 1 () inhibiting the cellular metabolism/' and (5) redistributing the lymphocytes.' The purpose of our investigation was to study the effect of locally administered corticosteroids on the immunologic response of AFC in the cornea, uveal tract, and draining lymph nodes after the intracorneal injection of a protein antigen (BGC). Materials and methods The: experiment;!! animals were New Zealand white: adult male: rabbits weighing four to five pounds. Prior to the experiment they had no ocular pathology and no palpable pre-auricular or cervical lymph nodes. In the first experiment, each of 2 rabbits was bled from the marginal ear vein before injection and again immediately 1 before being killed. The serum was separated I mm each specimen and inactivated by being heated at 5o' C. tor 3 minutes. It was then absorbed with an equal volume of washed sheep red blood cells and stored at -2 C. After placing two drops of propaiacainc hydrochloride (Ophthaine) in the conjunctival sac ol each animal's right eye, we injected intrastromally approximately.1 ml. BCC with a 3-gauge needle attached to a. ml. tuberculin syringe. The BCC (Miles Laboratories, Kankakee, III.) had been dissolved in.85 per cent sodium chloride solution and filtered through a Millipore filter of.5 M median pore size. The concentration of the solution was 1 nig. per millilitcr; the intracorneal close was approximately 1 mg. Immediately after the intracorneal injection ol BCC, the right eyes of 1 animals received a subconjunctival injection of. ml. ti iamcinolone acetonide suspension. The injection was given with a 27-gauge needle attached to a 1. ml. tuberculin syringe. The concentration of the triamcinolone suspension ( K'enalog, E. H. Squibb and Sons, New York, N. Y.) was mg. per milliliter; the subconjunctival dose was therefore approximately 1 ing. The same eye received a second 1 mg. close seven days later. A control group of 1 animals received subeonjiinctival injections of. ml. sterile saline solution, the first one immediately after the intracmneal injection of BCC. and the second seven days later. The rabbits were examined daily with the slit lamp. The degree of circumlimbal injection, corneal vascularization, corneal opacity, and iritis were graded by a pre-arranged scale of to 3 + which could be applied in a double-blind manner. Four rabbits (two that had received the cortieosteroid and two controls) were killed on each of clays fi,,,, and after the intracorneal injection of BCC. The spleen, the preauricular and cervical lymph nodes on the homolateral and contralateral sides, the homolateral uveal tract, and the homolateral cornea (including a 1 mm. rim of sclera) were removed and placed in cold Eagle's minimum essential medium (MEM, Hylancl Laboratories, Los Angeles, Calif.). Samples of aqueous and vitreous were removed and stored at -2 C. The corneal tissue was finely minced; lymph node, spleen, and uveal tissues were gently teased in dissecting dishes, and the resultant cell suspensions from each tissue were separately filteied through sterile fine-gauge nylon mesh to remove any large particles. All procedures were carried out in cold MEM at C. The hemolysis-in-agar technique of )erne, modified for use with protein antigens," was used to detect AFC in the lymph nodes, uveal tract, cornea, and spleen. Each cell suspension was plated in triplicate with sheep erytlirocytes, both BCC-coatcd and uncoated. The results of the placque assays were expressed as the number of placques per million cells in the original suspension. The average number of background plaeques on the control plates containing uncoated erythrocytes was subtracted from the number on the BCC-coated test plate. The serum, aqueous, and vitreous were tested tor hemolytic antibody by a modification of the technique described by Mubner and Cengozian." Twofold dilutions of each lluid were made with the micmtiter apparatus. A drop of each dilution was placed on the surface ol a layer of agarose that contained BCC-coated erythrocytes. Control plates contained uncoated erythrocytes. After the samples had been absorbed by the agarose, the plates were rinsed with MEM. Cuinea pig complement (1:1 dilution) was added and the plates were incubated for two hours at 37 C. A small area of lysis at the site where a drop of the test lluid had been placed on the plate indicated the presence of antibody. In a second experiment, we used rabbits to confirm the results of the first experiment and to accumulate more data. The procedure was the same as in the first experiment with the following exceptions: (1) animals were not killed on postinjection days fi and, and (2) the contralateral lymph nodes and spleens were not assayed for AFC. Wedge-shaped sections were taken from representative corneas in the steroid-treated and control groups on postinjection days,, and for histologic study. The sections were fixed in formalin and stained, some with hematoxylin and eosin, others with methyl green-pyronine. Downloaded from iovs.arvojournals.org on 1/26/

3 1 Meyer et al. hivvstiuntivc Oplithalinologij February 175 Table I. Clinical response of rabbits inoculated intracomeally with bovine gamma-globulin Postinjection Treatment daij Corticosteroids 6 y Normal saline f> solution (control) Rabbit No fi <S IS H Circumlimbal injection li Conical oa.sciilarizat.ion (.) (J Conical o pacification (.) Iritis Results In summarizing our results for the purposes of this report, we added together the 2 rabbits of Experiment 1 and the rabbits of Experiment 2. Accordingly, the clinical responses of the two groups inoculated intracorneally with BGG ( rabbits) appear in Table I, and the results of our antibody studies of the two groups ( rabbits) appear in Tables II and III. The biphasic response of the rabbit eye to an intracorneal injection of BGG has been described previously. 1 Throughout all stages of this biphasic response, the corticosteroid-treated eyes in the present experiment were less inflamed than the control eyes. Only one corticosteroid-treated eye developed circumlimbal injection, one developed mild corneal vascularization, and none showed corneal infiltrates or iritis. In contrast, nearly every control eye had some degree of conjunctival injection, corneal vascularization and infiltration, and iritis. The signs of inflammation were most severe on postinjection days and. Table II presents the results of placque assays on the lymph nodes, cornea, uveal tract, and spleen. The homolaleral lymph nodes contained AFC in approximately equal numbers in the corticosteroid-treated and control groups from postinjection day 6 through day. As noted in previous experiments, 1 the lymph nodes draining the uninjected left eyes did not contain AFC (exception: a few were found in the left lymph node of control rabbit No. 6). AFC were detected in the cornea from day through day in 1 (83 per cent) Downloaded from iovs.arvojournals.org on 1/26/

4 VUIIIIIW \ Nuinhcr 2 Effect of local corticosteroids 11 Table II. Antibody formation by cells ot the lymph nodes, cornea, uveal tract, and spleen of rabbits injected intraeorneally with bovine gammaglobulin Treatment Cortieosteroids Normal saline solution (control) Postinjection day 6 Rabbit No Ri-ht lymph node , , Tissue assayed: PUiajucs/1" cells' 1 Left lymph node Cornea , Uoeal tract , ,5 26 Spleen 1 7 "Minus li;k'k^roiiii<l-plikc ik' count. of the control eyes lmt in only one (S per cent) of the corticosteroid-treated eyes. There was a close correlation between the number of AFC in the cornea and the severity of the clinical corneal reactions. For example, none of the corticosteroid-treated eyes showed corneal opacification and only one had a few AFC, whereas 11 of the control eyes showed opacification and nine of the 11 had many AFC. The uveal tract contained AFC from postinjection day through day in nine (75 per cent) of the control eyes but in only three ( per cent) of the corticosteroid-treated eyes. The number of AFC in the uveal tract correlated well with the severity of the iritis. In only four spleens of the rabbits (two from corticosteroid-treated animals and two from control animals) could AFC be detected. The number of AFC in each of the spleens was small compared with the number in the corresponding draining lymph node. Table III presents the anti-bcc hemolytic antibody liters in the aqueous humor, vitreous humor, and serum of all rabbits. None of the eyes showed detectable antibody in the vitreous, and only three had detectable antibody in the aqueous, all in the control group. Circulating antibodies were present in the sera from postinjection day 6 through day, but there was no statistical dififere nee in their level Downloaded from iovs.arvojournals.org on 1/26/

5 12 Meyer et al. Investigative Ophthalmology February 175 Table III. Titeis (A hemolytic antibody present in aqueous humor, vitreous humor, and.serum of rabbits inoculated intracorneally with bovine gamma-globulin Postinjection Treatment daij Corticosteroids 6 Normal saline solution 6 (control) Rabbit No H IS AIJIICOIIS humor (.) 1:2 Vitreous humor Serum 1:6 1: 1: 1:2 1:16 1:6 1:6 1:16 1:6 1:6 between the corticosteroid-treated and control groups. There also appeared to be no direct correlation between the antibody titers and the severity of the ocular clinical reactions, or between the antibody titers and the number of AFC in the tissues. The histology correlated well with the clinical observations. In the control eyes with ring opacities, blood vessels extended centrally from the limbus and were accompanied by polymorphonuclear and mononuclear cells. The cellular reaction was greatest on postinjection days and. The corticosteroid-treated eyes showed qualitatively similar but relatively minor reactions. Polymorphonuclear and mononuclear cells were seen after the same time intervals whether the eyes were corticosteroid-treated or controls. Plasma cells were present in the corticosteroid-treated corneas that showed AFC but in only small numbers. Corneal vascularization was minimal in the corticosteroid-treatcd eyes. Discussion Corticosteroids are the drugs used most often to suppress undesirable immunologic responses. Much descriptive detail has been recorded regarding their effect, but little has been determined about the exact mechanisms by which they achieve their anti-immunologic and anti-inflammatory benefits. They may intervene at various points in the immune response and may affect many aspects of inflammation. Downloaded from iovs.arvojournals.org on 1/26/

6 Volume I'I Nu in her 2 Effect of local corticosteroids 13 In the present experiments, the escape of antigen from the cornea, followed by the formation of AFC in the draining lymph nodes, apparently proceeded normally. The question arises as to whether or not the corticosteroids actually reached the lymph nodes in time to prevent sensitization or if they reached the lymph nodes at all. This is a difficult problem to evaluate, however, it has been shown that intralamellar injection of India ink into vaseularized rabbit corneas is followed by rapid spread of the material to the limbus and ipsilateral regional lymph nodes. 1 " Also, subconjunctival '"I-albumin-Evans blue complex reaches the lymph nodes of the neck in rats within six minutes. When the Evans blue complex is injected into normal rat cornea, the time taken to reach the lymph nodes is about six hours." The cortices re roid molecule is water soluble and smaller than either the India ink or the Evans blue complex. From this experimental evidence, we feel that the corticosteroids probably reached the draining lymph nodes and arrived before the antigen which was injected into an avascular cornea. However, the corticosteroid dosage may have been insufficient to prevent sensitization. The suggestion that corticosteroids impair the transport of antigens from the graft to the regional node'-' was not substantiated by our results. The circulating antibody that appeared was probably produced by these lymph node cells and was unaffected by the corticosteroids. In the corticosteroid-treated eyes, however, AFC did not reach the corneas or uveal tracts in the same numbers that they did in the control eyes. It appears that AFC production occurs in the usual way but that the ability of the AFC to reach the cornea and uveal tract is impaired. Lymphocytolysis could play a role here but probably does not do so since corticosteroids do not lyse mature AFC. S Redistribution of the lymphocytes' 1 ' " may be the crucial factor since AFC are present mainly in the draining nodes; and corticosteroids may influence the release of AFC from the nodes, thereby preventing them from reaching the target tissues. The anti-inflammatory effect of locally administered corticosteroids appears to have been important in producing the results of our experiment. Certainly there was a marked difference in the inflammatory response between the steroid-treated and control groups of eyes. The corticosteroid-treated eyes had notably less blood vessel engorgement and neovascularization than the control eyes; and this may be a lactor since the resultant difference in blood flow to the area may reduce the number of AFC that reach the target tissue. In this way the corticosteroids may suppress an established corneal response but not its initiation. In experiments now in progress, we shall attempt to assess the effect of presentment with locally as well as systemically administered corticosteroids before the intracorneal inoculation of BCG, and to assess the effect of other anti-inflammatory agents on the cellular components of corneal immunologic reactions. REFERENCES 1. Smolin, C, and Hall,.: Alferent arc of the eomeal immunologic reaction. II. Local and systemic response to bovine gamma-globulin, Arch. Ophthalmol. : 231, Smolin, C, Hall, J., and Cignetti, F.: The allcrent arc of the cornea! immiinologic reaction: migration inhibitor)- factor, INVEST. OPHTHALMOL. : 2, Znrier, R. B., and Weissman, C: Antiimmunologie and anti-inflammatory ellects of steroid therapy, Med. Clin. N. Anier. 57: 5, VVeissman, C: The role of lysosomes in inflammation and disease, Ann. Rev. Med. 18: 7, Claman, H. N.: Corticosteroids and lymplioid cells, N ;. Engl.. Med. 287: 388, 172. fi. Colub, E. S., Mishell, R. I., Weigle, W. O., et al.: A modification of the hemolytic placqne assay for use with protein antigens, J. lmimuiol. 1: 133, Hubner, K. F., and Gengozian, VV.: Critical variables of the Jeme placquc technique as applied to rodent antibody-forming systems responding to heterologous red cell antigens, J. Immunol. : 5, 16. Downloaded from iovs.arvojournals.org on 1/26/

7 1 Meyer et al. Invfxtillative Fvhmnnj Cohen, J. J.: The elfects of hydrocortisone on the imimme response, Ann. Allcrg. 2: 8, Cohen, J. J.: Thymtis-derivcd lymphocytes sequestered in the hone marrow of hydrocortisone-treated mice, J. Immunol. 17: 81, Smolin, C, and Myndiuk, R. A.: Lymphatic drainage from vascularized rabhit cornea, Am. J. Ophthalmol. 72: 17, 171. : I- 11. Collin, B. H.: Lymphatic drainage of albumin from the vascularized cornea, IN- VEST. Ol'IITIIALMOL. : 16, 17.. Scothorne, H. ).: The ellect of cortisone on the cellular changes in the regional lymph node draining a skin homograft, Transplant. Bull. 3: 13, 156. Downloaded from iovs.arvojournals.org on 1/26/