Conjunctival epithelial cells cultured on human amniotic membrane fail to transdifferentiate into corneal epithelial-type cells.

PURPOSE: Previous studies on the use of human amniotic membrane (HAM) in rabbit stem cell deficiency models have found the new epithelium growing over the HAM to express cornea-specific keratins (K3 and K12) in 40% of the cases, suggesting that HAM may have induced conjunctival epithelial cells to transdifferentiate into cornea-type epithelial cells. The current study was performed to determine whether HAM could induce transdifferentiation of conjunctival epithelia] cells when cultured in vitro. METHODS: Conjunctival grafts taken from the fornices of New Zealand white rabbits (6-12 weeks old) were placed over HAMs and lifted to an air-media interface using polypropylene double rings. These cultures were maintained in supplemented hormonal epithelial medium with and without 3T3 feeder cells. Rabbit corneal epithelial cells were cultured similarly using strips of keratolimbal grafts placed over HAM. The cultures were terminated at various times between the 8th and 15th day. The cultured epithelial cells were examined histologically and immunohistochemically using monoclonal antibodies AK-2 (to K12 keratin), AM-3 (to goblet cell mucin), and AE-5 (to K3 keratin). RESULTS: Both conjunctival and corneal epithelial cells cultured on HAMs showed multilayered, differentiated epithelial structures. On immunohistochemical examination, both epithelial cells stained positive for AE-5. None of the cultured conjunctival epithelial cells stained positively for AK-2, while the corneal epithelial cells showed positive staining with AK-2. There were no AM-3-positive goblet cells in either epithelial cell culture. There was no difference in the immunohistochemical patterns between cultures with or without 3T3 feeder cells. However, culture without feeder cells seemed to manifest a more degenerative appearance than those with feeders. CONCLUSION: HAM does not induce transdifferentiation of conjunctival epithelial cells into corneal-type epithelial cells under the in vitro culture conditions used in this study.

Ex vivo transduction of corneal epithelial progenitor cells using a retroviral vector.

PURPOSE: To transduce corneal epithelial progenitor cells with a reporter gene using a retroviral vector and follow their progeny in vitro and in vivo. METHODS: Using a lacZ-producing retroviral vector, rabbit keratolimbal explants were transduced ex vivo, autografted onto their original sites, and assessed for lacZ-producing cells in the cornea throughout a 6-month period. Four autografts served as control samples, having received no vector. Experimental and control rabbits were euthanized and corneas with scleral rims harvested, weekly for 4 weeks and then monthly for 6 months. The corneas were first stained with 5-bromo-4-chloro-3-indolyl galactopyranoside (X-gal) as wholemounts and then sectioned for histology and immunohistochemistry to examine lacZ-positive cell outgrowth. Three additional transduced explants were observed in culture. These explants were transferred to new culture dishes every week for 9 weeks. The previously occupied culture dish was stained for lacZ to detect transduced epithelial cells, and the number of lacZ-positive cells was quantitated. RESULTS: LacZ-positive cells were found in the corneas of 18 of 20 eyes in which virally transduced keratolimbal autografts had been implanted. The cells were epithelial in nature, originated from the limbus, and were found in colonies throughout the epithelial layer of the cornea. The appearance of lacZ-positive cells in four of five corneas harvested after 6 months showed long-term transgene expression consistent with transduction of corneal epithelial stem cells. In vitro, the number of lacZ-positive cells migrating from the keratolimbal autografts decreased rapidly during the first 4 weeks and then remained stable through week 9. CONCLUSIONS: This study shows that a retroviral vector can effectively transduce corneal epithelial progenitor cells, shown by the long-term appearance of transduced cells on the cornea in vivo and the stable production of lacZ-positive cells in vitro. The appearance and disappearance of labeled cells is consistent with the initial transduction of stem cells and transient amplifying cells.

A new, simple, nonradioactive, nontoxic in vitro assay to monitor corneal endothelial cell viability.

PURPOSE: This study was designed to determine whether Alamar blue could be used to evaluate corneal endothelial cell viability in vitro. METHODS: Alamar blue incorporates a proprietary redox indicator that changes color in response to metabolic activity. Primary rabbit endothelial cells were subcultured on 96-well plates at densities ranging from 1,250 to 40,000 cells per well. After 12 hours' incubation, Alamar blue was added to each well and absorbance measured hourly from 1 to 9 hours. Sodium azide-killed cells were used as a control. Alamar blue conversion was also compared with [3H]-thymidine incorporation in the presence or the absence of mitomycin C. RESULTS: Alamar blue reduction demonstrated endothelial cell viability at all cell concentrations compared with that in killed-cell controls. The reduction varied proportionately with cell number and time, showing clearly significant differences. Conversely, [3H]-thymidine uptake demonstrated minimal DNA synthesis and little or no ability to distinguish cell number or viahility. CONCLUSIONS: Alamar blue reduction measures endothelial cell viability and can readily differentiate cell concentrations. It demonstrates several advantages over [3H]-thymidine: It can assay nonproliferating endothelial cell metabolism, it allows rapid assessment of large numbers of samples, it can differentiate endothelial cell concentrations, it is nontoxic, it is nonradioactive and allows for simple disposal, it is less costly, and it allows for continuous monitoring of endothelial cell metabolism and viability.

Povidone-iodine (betadine) in the treatment of experimental Pseudomonas aeruginosa keratitis.

Topical 5% povidone-iodine for the treatment of corneal ulcers was observed in Sierra Leone, West Africa by one of us (D.J.D.). To test the efficacy of topical 5% povidone-iodine for infectious keratitis, experimental Pseudomonas aeruginosa keratitis was induced in 12 rabbits by first abrading the central 3 mm of corneal epithelium. Thirty milliliters of broth of P. aeruginosa strain ATCC 27835 (1.8 x 10(7) viable bacteria) was dropped twice on the wounded cornea. After 22 h, all corneas were clinically infected. Eight rabbits were treated with 5% povidone-iodine solution and four with 0.9% NaCl solution. All were given hourly drops. Twenty-four hours after treatment began, the central 8-mm button of the infected cornea was excised, homogenized, and serial dilutions plated onto MacConkey agar. The total number of viable Pseudomonas organisms was calculated. The treatment group had 5.2 +/- 0.4 CFUs (colony-forming units) per cornea. The control group had 4.8 +/- 0.4 CFUs per cornea (p = 0.11). The clinical scores (Hobden grading system) were 6.9 +/- 1.5 for the treated group and 7.3 +/- 2.5 for the control group (p = 0.74). There was no statistical difference between the treated and control groups. Povidone-iodine (5%) is not effective in the acute treatment of P. aeruginosa keratitis in this rabbit model.

Oral tolerance in experimental autoimmune uveoretinitis. Distinct mechanisms of resistance are induced by low dose vs high dose feeding protocols.

Studies of oral tolerance in LEW rat models of autoimmune diseases including S-antigen (S-Ag)-mediated experimental autoimmune uveoretinitis (EAU), and myelin basic protein-induced experimental autoimmune encephalomyelitis have produced conflicting evidence for the roles of clonal anergy and suppression. Using subpeptides from a region of S-Ag known to induce oral tolerance a protective site was localized to a nonamer of residues 347-355. This site was also uveitogenic, providing the basis for testable hypotheses for tolerance to be due to clonal anergy in pathogenic T cells specific for that site, or to suppression. Evidence for suppression was strongly supported by several observations. 1) Induction of oral tolerance with low dose feedings (250 micrograms/feeding) of peptide 343-362 conferred resistance to EAU induction by intact S-Ag, which should not be possible if only T cells specific for epitopes in 343-362 were rendered unresponsive, since there are several other pathogenic sites in S-Ag. 2) Low dose feeding induced resistance to EAU induction by a distinct, spatially separate peptide, residues 270-289, of S-Ag. 3) The requirement for linked recognition was shown by the inability of tolerance induced by feeding 343-362 to protect from EAU induction by a peptide, residues 521-540, derived from interphotoreceptor retinoid binding protein, a different uveitogenic retinal protein. 4) Resistance could be adoptively transferred. Conversely, induction of tolerance with high doses of peptide (5 mg/feeding) resulted in loss of resistance to EAU induced by S-Ag, although disease induction by the fed peptide was inhibited; observations that are consistent with clonal anergy. The apparent lack of suppression after high dose feeding could mean that suppressor T cells can also be rendered unresponsive or that induction of T suppressor cells is dependent on CD4+ Th cells, which were rendered anergic, leading to lack of T suppressor development. We suggest that oral tolerance operates by at least two distinct mechanisms that depend on the feeding dose; low doses induce suppression, whereas high doses induce unresponsiveness.

Inhibition of in vitro T cell activation by corneal endothelial cells.

Cells and tissues of the anterior uvea and aqueous humor express activities which inhibit immune responses. These activities include soluble factors such as TGF-beta and uncharacterized cell surface interactions. Relatively little is known regarding the immunologic activities of corneal endothelium, despite its potentially important role in contributing to the immune privilege of the anterior chamber and the high success rate of corneal transplantation. In this report, in vitro studies of cultured rat corneal endothelial (CE) cells were done using S-antigen-specific LEW rat T cell lines, or S-antigen-specific T cell hybridomas, to examine the immunologic capabilities of CE cells. Monolayers of LEW rat CE cells were unable to present antigen or a mitogen, Con A, to T cell lines or hybridomas as assessed by the lack of a proliferative response or IL-2 secretion. Furthermore, the CE cells exerted a potent inhibitory effect when added to in vitro proliferation assays of T cell lines stimulated with antigen or Con A. When T cells were preactivated on conventional antigen presenting cells and then transferred to wells containing CE cells, their proliferation was not inhibited. Although CE cells inhibited activation of T cell lines and hybridomas, they did not inhibit the growth of T cell hybridomas or CTLL cells, nor did the CE cells adversely affect the viability of resting T cells cultured on CE monolayers. The inhibitory effect was reversible as preincubation of T cells on CE cells for up to 6 days followed by washes restored T cell responsiveness when assayed on splenocytes. The inability to stimulate proliferative responses was not affected by preincubation of the CE cells with lymphokines which increase MHC antigen expression. The inhibition observed in these assays was not MHC-restricted as CE cells from both LEW and BN rats were equally inhibitory. CE cells from rabbits and cats were also potent inhibitors of T cell activation, suggesting that the mechanism is evolutionarily conserved. The mechanism of inhibition of CE cells is unknown at this time.

Identification of a potent new pathogenic site in human retinal S-antigen which induces experimental autoimmune uveoretinitis in LEW rats.

Experimental autoimmune uveoretinitis (EAU) is a T cell-mediated autoimmune disease of the eye which can be induced in LEW rats by immunization with either human or bovine S-antigen (S-Ag). In previous reports, two nonimmunodominant pathogenic sites were found using synthetic peptides corresponding to conserved sequences at amino acid residues 303-314 and 286-297 of the bovine sequence. In this report, a 20-residue synthetic peptide encompassing amino acids 343-362 located near the C-terminus was found to be highly immunopathogenic in LEW rats. The onset of EAU was observed at as early as 8 days when high doses of a peptide-encompassing residues 343-362 were used. EAU was elicited with as little as 0.5 microgram of peptide per animal. Smaller peptides from this region were also tested for uveitogenicity, further refining the site to 13 amino acids. Uveitogenic T cell lines were made to this site in two ways; first, by the in vitro selection of a bulk T cell line raised to human S-Ag with peptide 343-362. Second, by the in vitro selection of a peptide-specific line from an animal immunized with peptide 352-364, which corresponds to the minimal uveitogenic site. Both of these lines adoptively transferred EAU to LEW rats, further establishing the pathogenicity of this site. A proliferative site distinct from, but overlapping, the uveitogenic site was also found. The potent uveitopathogenicity of peptides from this region indicates that it is a major pathogenic site responsible for EAU induced in LEW rats by immunization with human S-Ag.

Multiple, spatially distinct T cell epitopes within a pathogenic 123 residue cyanogen bromide peptide of bovine retinal s-antigen.

We have examined the T cell specificity of a Lewis rat T cell line (R208) specific for a pathogenic, 123 residue cyanogen bromide produced peptide of bovine S-antigen by using two independent sets of overlapping synthetic peptides representing the entire length of the 123 residue fragment. S-antigen, a 48 kDa immunopathogenic photoreceptor cell autoantigen induces T cell mediated experimental autoimmune uveoretinitis (EAU) in experimental animals. Extensive analyses revealed a heterogenous response by the R208 line to the panel of synthetic peptides, proliferating weakly to 4 distinct sites. Unexpectedly, peptides representing sequences (residues 286-297 and 303-320 of bovine S-antigen) known to actively induce the autoimmune pathology were unable to significantly stimulate the R208 line as assessed by proliferation assays. Similarly, attempts to isolate T cells specific for these sequences from the R208 line have proven unsuccessful. However, two sequences, residues 253-269 and 273-289, sufficiently stimulated R208 cells to allow isolation of sub-lines, R208:26 and R208:28, respectively. Neither of these peptides actively induce an autoimmune response. R208:26 does not transfer EAU and R208:28 transfers moderate EAU. As a control, we are able to isolate a pathogenic T cell line (R502) specific for the actively pathogenic sequence, residues 303-320, when this peptide is used as the immunogen. However, the R502 line proliferates to peptides (e.g. 305-322) which do not contain residues 303 and 304 which are critical for the active induction of disease. These results show a multiplicity of distinct T cell epitopes within a relatively small region of S-antigen.(ABSTRACT TRUNCATED AT 250 WORDS)

A new perspective of S-antigen from immunochemical analysis.

Elucidation of the amino acid sequences of retinal S-antigens from several species has allowed the fine dissection of T cell and antibody epitopes using synthetic peptides. S-antigen, isolated from retinal rod photoreceptor cells, elicits experimental autoimmune uveoretinitis (EAU), a predominantly CD4+ T-cell mediated autoimmune disease of the retina and uveal tract of the eye and pineal gland. Three uveitogenic T cell lines, R9, R17 and R208, prepared against native bovine S-antigen, human S-antigen and cyanogen bromide peptide CB123, respectively, were used to identify the T cell recognition sites responsible for uveitogenic and proliferative responses. T cell epitopes were found to be clustered into 6 regions, some of which were species-specific. The two synthetic peptides known to actively induce EAU, residues 286-297 and 303-314 of bovine S-antigen, were unable to induce significant proliferative responses in any of the three T cell lines. However, both of these sites were adjacent to synthetic peptides, residues 273-292 and 317-328, respectively, which were unable to actively induce EAU, but elicited proliferative responses from the T cell lines. We also report the presence of a new pathogenic site, also associated with an adjacent proliferative site, together in residues 343-362 of bovine S-Ag. Our results indicate that spatially separate and distinct T cell epitopes are present in S-antigen which are responsible for the active induction of EAU, lymphocyte proliferation, and adoptive transfer of EAU.

Identification of T cell recognition sites in S-antigen: dissociation of proliferative and pathogenic sites.

Experimental autoimmune uveoretinitis (EAU) is a predominantly CD4+ T cell-mediated autoimmune inflammatory disease of the retina and uveal tract of the eye and the pineal gland. S-antigen, a protein found in retinal photoreceptor cells and pinealocytes, is a potent agent for the induction of EAU in susceptible species and strains. In order to identify the T cell recognition sites of S-antigen responsible for its uveitogenicity and proliferative responses, cyanogen bromide (CB) fragments as well as synthetic peptides were used to test the proliferative responses of two uveitogenic T cell lines, R9 and R17, prepared against native bovine and human S-antigen, respectively. Two nonoverlapping synthetic peptides which are known to actively induce EAU, amino acid residues 286-297 and 303-314 of the bovine sequence, were unable to induce proliferative responses in either S-antigen-specific T cell line. However, both of these sites were adjacent to synthetic peptides, residues 273-292 and 317-328, respectively, which were unable to actively induce EAU, but elicited strong proliferative responses from T cell lines raised to bovine and human S-antigen. Repeated in vitro selection of the R9 T cell line with a synthetic peptide containing one of these proliferative sites, residues 317-328, gave rise to a transiently uveitogenic T cell line. Several species-specific T cell epitopes were identified, but none of these were found to be involved in a uveitogenic response. Our results indicate that spatially separated and distinct T cell epitopes are present in S-antigen which are responsible for the active induction of EAU, lymphocyte proliferation, and the ability to adoptively transfer EAU.

Identification of a uveitogenic cyanogen bromide peptide of bovine retinal S-antigen and preparation of a uveitogenic, peptide-specific T cell line.

Peptide fragments of bovine retinal S-antigen produced by cyanogen bromide (CB) digestion have been purified and tested for their ability to induce experimental autoimmune uveoretinitis (EAU) and pinealitis (EAP) in Lewis rats. Following immunization with the various peptides in complete Freund's adjuvant, one of the peptides, CB123, was found to be potently uveitogenic. A CB123-specific, class II restricted T helper lymphocyte line (R208) prepared from one of the CB123 peptide-immunized animals by repeated in vitro selection with purified CB123 was able to transfer severe EAU and EAP to naive rats. Peptides CB36, CB46, CB51 and CB66 were immunogenic as assessed by the presence of antibodies detected in the enzyme-linked immunosorbent assay and sensitized lymphocytes found in [3H]thymidine incorporation assays using lymphocytes from the peptide-immunized rats, but they did not induce significant EAU or EAP. Antibodies induced by immunization with the peptides also bound intact S-antigen, unlike the lymphocytes which were only weakly responsive to S-antigen. Cross-reactivity of antibodies and lymphocytes from CB51 and CB66 immune animals suggested that these peptides overlap and contain a common epitope. No measurable immunity to any CB peptide or intact S-antigen was found in the animals immunized with peptide CB26. Although a T cell line (R17) raised to human S-antigen was also able to transfer EAU, it was weakly responsive to the CB123 peptide; also, the R208 line did not respond well to human S-antigen. Since both human and bovine S-antigen are uveitogenic, these results suggest that the uveitogenic epitope in CB123, which is of bovine origin, is not the same as the uveitogenic epitope of the human antigen.

S-antigen-specific rat T cell lines recognize peptide fragments of S-antigen and mediate experimental autoimmune uveoretinitis and pinealitis.

Two S-antigen-specific rat T cell lines expressing the T helper cell surface phenotype (W 3/25+, OX 8-) have been isolated from the spleen and lymph node cells of retinal S-antigen-immunized Lewis rats, one of which displayed neither clinical nor histopathologic signs of experimental autoimmune uveoretinitis. The other rat had recovered from severe experimental autoimmune uveoretinitis for 2 mo before isolation of the cell line. Both lines are specific for S-antigen presented by histocompatible antigen-presenting cells, and also respond in vitro to several of the peptides produced by cyanogen bromide cleavage of bovine retinal S-antigen. The lesions induced by the i.v. transfer of from 1 to 10 X 10(6) viable line cells involve the retina and pineal gland, as is found when Lewis rats are immunized with immunopathogenic doses of S-antigen. Histologic examination of the eyes and pineal glands revealed pathologic lesions typical of experimental autoimmune uveoretinitis, and consisted of marked infiltration of the retina and surrounding tissues and the pineal gland by lymphocytes and inflammatory cells. T cells capable of mediating autoimmune disease are clearly present and readily isolated from both asymptomatic and convalescent animals. No significant differences in specificity for the cyanogen bromide peptides of S-antigen or cell surface phenotype were found in the T cell lines isolated from these two rats, nor was any difference found in the specificity or titer of serum antibodies taken from the original rats for the cyanogen bromide peptides of S-antigen.