IOVS, May 1999, Vol. 40, No. 6 Reports 1275

fore not clear whether the vascular regulation might be equally 3. Graham SL, Drance SM, Wijsman CJ, Mikelberg FS, Douglas GR. impaired in these two situations. This observation suggests that Ambulator)' blood pressure monitoring in glaucoma patients: the the possible vascular dysfunction underlying the glaucomatous nocturnal clip. Ophthalmology. 1995; 102:61-69. process is not tightly related to the stage of the disease. 4. Harris A, Sergott RC, Spaeth GL, et al. Color Doppler analysis of ocular vessel blood velocity in normal tension glaucoma. Am / Topicical dorzolamide reduced the IOP in our patients with Ophthalmol, 1994;118:642- 649. glaucoma (groups B and C). Our measurements showed that 5. Lieb WE, Cohen SM, Merton DA, ShieldsJA, Mitchell DC, Goldberg both, initial and advanced glaucomatous eyes, have lower IOPs BB. Color Doppler imaging of the eye and orbit: technique and (195 ± 3-6 and 19.6 ±31 mm Hg, respectively) after dorzol- normal vascular anatomy. Arch Ophtbalmol, 1991; 109:527-531. amide application than under baseline conditions (24.2 ± 5.3 and 6. Sergott RC, Aburn NS, Trible JR, Lieb WE, Falaharty PM. Color 24.1 ±4.1, respectively). Moreover, we have also observed that Doppler imaging: methodology and preliminary results in glau- the IOP increased in groups B and C after the initial dorzolamide coma. Sum Ophthalmol. l994;38(suppl):S65-S71. withdrawal. Our observation that dorzolamide reduces the IOP in 7. Williamson TH, Harris A. Color Doppler ultrasound imaging of the glaucomatous eyes and that it shows an additive effect with eye and orbit. Surv Ophthalmol. 1996;40:255-267. j3-blockers fully agrees with the observations reported in the 8. Kaiser HJ, Schoetzau A, Stiimpfig D, Flammer J. Blood flow veloc- 17 ities of the extraocular vessels in patients with high-tension and literature. On the other hand, in our study this hypotensive normal-tension primary open-angle glaucoma. Am J Ophthalmol. effect was not observed in healthy eyes. 1997;123:320-327. In summary, the main finding in the present study was that 9. Rankin SJA, Walman BE, Buckley AR, Drance SM. Color Doppler dorzolamide induces changes in ocular and periocular hemo- imaging and spectral analysis of the optic nerve vasculature in dynamics, improving blood perfusion of the eye. Improvement glaucoma. Am/ Ophthalmol, 1995;! 19:685-693- of blood flow, however, was not similar in all studied vessels. 10. Taylor KJW, Holland S. Doppler US, part 1: basic principles, instru- Hemodynamic changes in the OA were similar in normal and mentation and pitfalls. Radiology. 1990;l74:297-307. 11. Williamson TH, Baxter GM, Lowe GDO. The influence of age, glaucomatous eyes. In this vessel the PSV seems not to be systemic blood pressure, smoking and blood viscosity on orbital affected, whereas the EDV and the RI were improved. On the blood velocities. Br J Ophthalmol, 1995;79:17-22. contrary, dorzolamide induced different hemodynamic 12. Baxter GM, Williamson TH, McKillop G, Dutton GN. Color Dopp- changes in the CRA of normal and glaucomatous eyes. In this ler ultrasound of orbital and optic nerve blood flow: effects of vessel, the PSV increased in glaucomatous eyes but not in posture and timolol 0.5%. Invest Ophthalmol Vis Sci. 1992;33: normal eyes. The effect of the drug on the remaining hemody- 604-610. namic parameters of the CRA and CRV was similar in normal 13- Aim A, Bill A. The oxygen supply to the retina, II: effects of high and glaucomatous eyes. This differential effect of dorzolamide intraocular pressure and of increased arterial carbon dioxide ten- sion on uveal and retinal blood flow in cats—a study with labeled on the CRA of normal and glaucomatous eyes may be ex- microspheres including flow determinations in brain and some plained by the more prominent effect of the drug in lowering other tissues. Acta Physiol Scand. 1972;84:306-319. the IOP in the second group of eyes. 14. Harris A, Joos K, Kay M, Evans D, Shetty R, Sponsel WE. Acute IOP elevation with scleral suction: effects on retrobulbar haemodynam- Acknowledgments ics. Br J Ophthalmol. 1996;80:1055-1059- 15. Pillunat LE, Stodtmeister R, Wilmanns 1, Christ T. Autoregulation of The authors thank Jose Coclesido and Jose L. Otero for their help with ocular blood flow during changes in intraocular pressure, prelim- the color Doppler and the statistics. inary results. Graefes Arch Clin Exp Ophthalmol, 1985;223:219- References 223. 16. Harris A, Arend O, Arend S, Martin B. Effects of topical dorzol- 1. Drance SM. Glaucoma: a look beyond intraocular pressure. Am J amide on retinal and retrobulbar hemodynamics. Acta Ophthal- Ophtbalmol. 1997;123:817-819. mol Scand. 1996J 4:569-572. 2. Tielsch JM, KatzJ, Sommer A, Quigley HA, JavittJC. Hypertension, 17. Lippa EA. Carbonic anhydrase inhibitors. In: Rich R, Shields MB, perfusion pressure, and primary open-angle glaucoma: a popula- Krupin T, eds. TJie Glaucomas. New York: Mosby; 1996:1463- tion-based assessment. Arch Ophthalmol. 1995;113:2l6-221. 1481.

Epitope Mapping of Anti- the characteristics, origin, and possible significance of these antibodies, the epitopic specificity of the anti-rho- Rhodopsin Antibodies from dopsin antibodies was examined in four NPG patients. Patients with Normal METHODS. Antibodies in patient sera were assayed by west- Pressure Glaucoma ern blot analysis against purified bovine rhoclopsin. Pep- tides derived from particular segments of the rhodopsin Carmelo Romano,l Zhengzhi Li,x sequence were tested for activity in competing for rho- Anatol Arendt,2 Paul A. Hargrave,2 and dopsin-antibody binding. Martin B. Wax1 RESULTS. Of a series of nine peptides that constitute most of the hydrophilic regions of rhodopsin, only one, consist- PURPOSE. The presence of anti-rhodopsin antibodies in ing of the C-terminal 25 amino acids, prevented binding of patients with normal pressure glaucoma (NPG) has been the patient antibodies to rhodopsin. Higher resolution previously demonstrated with western blot analysis and mapping using a set of dodecamers of overlapping se- enzyme-linked immunosorbent assay. To learn more about quences from the C-terminal region demonstrated that

Downloaded from iovs.arvojournals.org on 10/01/2021 1276 Reports IOVS, May 1999, Vol. 40, No. 6

antibody binding is completely dependent on the last two Among the most interesting and perplexing associations amino acids. Removing the C-terminal alanine alone, or of glaucoma and autoimmunity are our previous findings that amidating the C terminus carboxyl group, also eliminated autoantibodies to rhodopsin are elevated in patients with nor- antibody binding. mal pressure compared with those with high pressure glau- coma or age-matched control patients.'' Glaucoma is tradition- CONCLUSIONS. Because four of four patient antibodies ex- ally considered to be a disease of ganglion cell layer loss, with amined exhibited the identical epitopic specificity, it is little or no subjective symptoms or signs that suggest photore- likely that a common mechanism underlies their genera- ceptor involvement. Therefore, although we considered the tion. This may indicate that molecular mimicry has oc- presence of elevated serum antibodies to rhodopsin a biochem- curred, because several pathogens contain similar C-ter- ical parameter indicating that the mechanism(s) of optic atro- minal sequences. Although they may serve as diagnostic phy in NPG and high pressure glaucoma may be distinctly markers, and provide evidence that there is an autoim- different, we were unable to conclude that elevated rhodopsin mune component in some patients with glaucoma, the antibodies were of pathogenic importance. To better under- role, if any, that these antibodies play in the pathogenesis stand the origin and potential significance of anti-rhodopsin of the disease remains unknown. (Invest Ophthalmol Vis antibodies in patients with NPG, we performed epitope map- Sci. 1999;40:1275-1280) ping to define in detail the immunogenic specificity of these antibodies. laucoma is no longer viewed simply as elevated intraocu- Glar pressure that damages the optic nerve. Studies have METHODS indicated that in a significant percentage of patients with glau- coma, intraocular pressure has never been demonstrated to be Patients elevated. The prevalence of this form of the disease, often All subjects were treated in accordance with the tenets of the called "normal pressure" or "low tension" glaucoma, is conser- Declaration of Helsinki. Blood samples were taken after de- vatively estimated to be approximately 20% to 30% of patients tailed consent was obtained from patients with NPG. Serum >2 with glaucoma.' Apoptotic cell death has been proposed to was prepared, heat inactivated, and stored at — 70°C until use. mediate the glaucomatous process in both high and normal The inclusion and exclusion criteria for these patients were pressure forms of experimental and human glaucoma.3"6 Al- described previously.1112 Briefly, NPG consisted of the pres- though neurodegenerative processes such as ischemia, excito- ence of open iridocorneal angles, no evidence of intraocular toxicity, neurotrophin insufficiency, and peroxynitrite damage pressure greater than 23 mm Hg, glaucomatous changes in have all been hypothesized to play a role in causing apoptotic visual fields and optic nerve cupping, and the absence of glaucomatous ganglion cell loss, none of these mechanisms alternative causes of optic neuropathy. Visual field loss of have been definitively linked to either the high or normal patients was evaluated with the Humphrey Field Analyzer, 30-2 pressure form of the disease. program (Humphrey Instruments, San Leandro, CA). Our cri- We have proposed that one form of normal pressure teria for visual field abnormalities included a corrected pattern glaucoma (NPG) may represent an autoimmune neuropathy. SD with a P < 0.05 or a glaucoma hemifield test outside normal Evidence that supports this includes immunoglobulin deposi- limits obtained with at least two reliable and reproducible tion in the ganglion cell layer observed in the postmortem visual field examinations. examination of eyes from a patient with NPG,6 an epidemio- logic association of immune-related disease in 30% of patients Peptide Synthesis with NPG,7 and the presence of aberrant or elevated serum Solid-phase peptide synthesis was performed using phenylac- autoantibodies including monoclonal paraproteins; antibodies etamidomethyl resin using an Applied Protein Technology to DNA, RNA, and nuclear proteins; and antibodies to heat- peptide synthesizer (model Pss 80; Cambridge, MA). tert-Bu- shock proteins in patients with NPG.8'9 In addition, autoanti- tyloxycarbonyl amino acids were coupled using diisopropyl- bodies to small heat-shock proteins are elevated in patients carbodiimide/yv-hydroxybenzotriazole, and the completed peptide was cleaved from the resin using liquid hydrogen with glaucoma, and these antibodies induce apoptosis in native 13 human retina and immortalized retinal cells in culture.10 fluoride. Peptides were purified by preparative high-perfor- mance liquid chromatography and were demonstrated to have the expected composition by amino acid analysis. In some cases, peptides were obtained from commercial From the 'Department of Ophthalmology and Visual Sciences, sources (Research Genetics, Huntsville, AL). Washington University School of Medicine, St. Louis, Missouri; and the 2Department of Ophthalmology, University of Florida, Gainesville. Supported in part by Grants EY12314, 02687, 06225, and 06226 Western Blot Analysis from National Eye Institute, National Institutes of Health, Bethesda, Patient sera were routinely tested for reactivity toward bovine Maryland; the Glaucoma Research Foundation, San Francisco, Califor- rhodopsin because we had previously demonstrated that NPG nia; the American Health Assistance Foundation, Washington, DC; and patients have serum antibodies that recognize bovine rhodop- unrestricted departmental grants from Research to Prevent Blindness, 11 New York, New York. sin. There is 96% amino acid identity (99% identity or con- Submitted for publication October 8, 1998; revised December 16, servative substitution) between bovine and human rhodopsins. 1998; accepted January 20, 1999. Bovine rhodopsin was prepared by chromatography on Con A Proprietary interest category: N. Sepharose in dodecyl maltoside, based on the procedure of Reprint requests: Carmelo Romano, Department of Ophthalmol- 14 ogy and Visual Sciences, Washington University School of Medicine, Litman. Some experiments used rhodopsin generously pro- Campus Box 8096, 660 South Euclid Avenue, St. Louis, MO 63110. vided by Steven Fliesler (St. Louis University, St. Louis, MO).

Downloaded from iovs.arvojournals.org on 10/01/2021 IOVS, May 1999, Vol. 40, No. 6 Reports 1277

Each gel lane contained 1 jag or 2 /u,g rhodopsin (determined full length rhodopsin = 348 aa using the BCA assay; Sigma Chemicals, St. Louis, MO). • • : ! i j! Proteins were separated by electrophoresis in 12% sodium 1st set.peptides | ! I dodecyl sulfate-polyacrylamide gels (SDS-PAGE) and electro- p1 = 2-32 phoretically transferred to polyvinyl membranes (Millipore, p2 = 65-75 D Marlboro, MA) in a Bio-Rad MiniTrans Blot Apparatus (Bio- p3 = 96-115 d : I i • Rad, Hercules CA). After transfer, membranes were incubated p4 = 142-153 I 1 in TTBS (50 mM Tris-HCl, 154 mM NaCl, 0.1% Tween-20, pH D I J 7.5) containing 2.5% nonfat dry milk for at least 15 minutes to p5 = 176-202 • i r ! block nonspecific protein adsorption to the membrane. Strips p6 = 231-252 • corresponding to individual gel lanes were cut out and used for p7 = 276-286 D LII incubation with patient sera in the presence or absence of p8 = 310-321 competing peptides. For competition, peptides (50 jag/ml) DL were incubated with diluted antisera (usually 1:2000 final) for p9 = 324-348 1 hour before application to strips. Antibody containing solu- tions, including sodium azide (0.1% wt/vol) as a preservative, p9, C-term peptide 324-348 were added to the strips for overnight incubation. After several 9nri RPt peptiries washes in TTBS, the membranes were incubated in TTBS/2.5% sp1 = 323-334 milk containing goat anti-human fx chain coupled to horserad- sp2 = 325-336 ish peroxidase (2000/1; Sigma) for 2 hours. After several sp3 = 327-338 washes in TTBS, immunoreactive bands were visualized by enhanced chemiluminescence using commercial reagents sp4 = 329-340 (products from Amersham, Arlington Heights, IL, and NEN, sp5 = 331-342 Boston, MA, gave comparable results). sp6 = 333-344 sp7 = 335-346 sp8 = 337-348 RESULTS -1 = 337-347 Antibodies that bind to macromolecules are actually directed amide = sp8-NH2 to small structural features, or epitopes, that usually consist of small delimited regions of the primary sequence. Such antibod- FIGURE 1. Schematic diagram of peptides derived from the sequence ies will bind to small peptides containing the epitopic se- of bovine rhodopsin used as competitors for antibody binding in this quences. Therefore, to determine the epitopic specificity of study. antibodies directed against rhodopsin found in NPG patient sera, we incubated the sera with peptides derived from the rhodopsin sequence (Fig. 1) and determined whether the pep- rhodopsin, the first beginning at residue 323 (peptide spl, tides prevented interaction with full-length rhodopsin immo- amino acids 323-344) and the start of the following peptides bilized on a western blot. serially incremented by 2 amino acids until the actual C termi- This strategy is illustrated in Figure 2A, using two mouse nus was included (sp8, amino acids 337-348). monoclonal antibodies directed against different parts of the When these peptides were incubated with sera from pa- rhodopsin sequence. Antibodies K16-107 and B6-3O are di- tient MW, only sp8 provided competition (spl and sp 8 shown rected against the C- and N-terminal sequences of rhodopsin, in Fig. 3). These data indicate that the C-terminal two amino respectively.'3 Incubation of either antibody with its cognate acids (proline-alanine, PA) are critical components of the peptide completely prevented binding to rhodopsin on the epitope recognized by this antibody. Furthermore, all verte- western blot. This is not a nonspecific inhibitory effect of brate opsins end with the sequence PA. either peptide, because they were completely ineffective at Because the last two amino acids were necessary for preventing the binding of the disparate antibody. antibody interaction, we wanted to determine whether in fact Nine peptides (pl-p9, Fig. 1) spanning the hydrophilic the C-terminal amino acid was critical. Peptide "-1," which is segments of the rhodopsin sequence were synthesized and identical to peptide sp8 except that it lacks the C-terminal tested for competition against sera from patient MW, a patient alanine, was synthesized and tested. This peptide did not com- with anti-rhodopsin antibodies that provide a robust signal on pete for binding (Fig. 3B), indicating a requirement for the western blot analysis.'' As shown in Figure 2B, only p9, con- C-terminal alanine as a component of the epitope. sisting of the C-terminal 25 amino acids, prevented binding. Because the data indicated that the C-terminal residue of This is strong evidence that the antibody present in this serum rhodopsin was an important component of the epitope recog- is monospecific and directed toward the C-terminal region of nized by this antibody, we tested the possibility that the car- rhodopsin. boxyl group denning the C terminus may participate in the Because individual epitopes are much smaller than 25 rhodopsin-antibody interaction. The C-terminally amidated an- amino acids, the next experiment was aimed to determine alogue of sp8 was synthesized. This substitutes the uncharged which subset of residues of the C-terminal 25 were critical for amide for the negatively charged carboxyl. This peptide was antibody binding. A second set of peptides (spl-sp8, Fig. 1) unable to prevent antibody binding to rhodopsin (Fig- 3B), was synthesized. Each peptide was 12 amino acids in length as pointing to a critical role for the negatively charged C-terminal determined by sequences taken from the C-terminal region of carboxyl group in antibody binding.

Downloaded from iovs.arvojournals.org on 10/01/2021 1278 Reports IOVS, May 1999, Vol. 40, No. 6 A. B. peptide p1 p9 - p1 p9 - p1 p2 p3 p4 p5 p6 p7 p8 p9 tr --• I

mAb K16-107 B6-30 (anti-C term) (anti-N term) patient (MW) serum

FIGURE 2. Specific competition of antibody binding to rhodopsin on western blot analysis by appropriate peptides. (A) Control antibodies of known specificity. Preincubation of monoclonal antibodies (mAb) directed against the N-terminal (B6-30) or C-terminal (K16-107) region of rhodopsin without (—) or with peptides derived from the N-terminal (pi) or C-terminal (p9) region prevented antibody-rhodopsin interaction only for the antibody directed toward the relevant portion of the sequence. Arrow indicates position of the rhodopsin monomer band. (B) C-terminal peptide competes for patient antibody-rhodopsin interaction. Serum from patient MW was preincubated with peptides of sequence derived from the hydrophilic portions of bovine rhodopsin. Only p9, containing the C-terminal 25 amino acids, prevented antibody-rhodopsin interaction,

We repeated all the experiments above using sera from for rhodopsin binding with each of the three sera (pi and p9 three additional patients. In our experience, human sera usu- illustrated in Fig. 4). Of the C-terminal 12 amino acid peptides, ally exhibit a high background staining, visible on the blots only sp8, the one containing the C terminus, was an effective shown in Figure 4. Nonetheless, the results of the peptide competitor of each of the three sera (spl, sp7, and sp8 illus- competition experiments were unambiguous and reproduc- trated in Fig. 4). These data indicate that, like the antibody ible. All three patients had anti-rhodopsin antibodies of identi- from patient MW, each of these three patient antibodies was cal epitopic specificity as patient MW (Fig. 4). monospecinc and directed toward the C terminus of rho- Of the initial set of 9 peptides that span the whole struc- dopsin. ture, only p9, containing the C terminus, effectively competed When peptide competition was performed using the pep- tide that lacked the C-terminal alanine, or the peptide that had the amidated C terminus, no competition was seen against any A. B. of the three patient antibodies (Fig. 4). Therefore, the anti- sp1sp8 . -1 am sp8 rhodopsin antibodies present in the sera of these three addi- tional NPG patients have an identical epitopic specificity to 113 antibody from patient MW. f DISCUSSION Our findings reveal that the serum autoantibodies in patients with NPG are directed toward the C terminus of the rhodopsin molecule. Because the competition by the C-terminal peptide was complete, it is highly unlikely that antibodies directed toward epitopes other than the C terminus contribute to the signal observed on the western blot analysis. Our results do not preclude the possibility that antibodies against conformation- FIGURE 3- (A) The last two amino acids of rhodopsin are critically ally dependent epitopes may also be present, but because our required for antibody binding. Eight small peptides derived from the C assay utilizes denatured rhodopsin they would not be detected. terminus (see Fig. 1 and text) were tested for their ability to compete The peptide competition experiments demonstrate the critical for patient antibody binding to rhodopsin. Only sp8, which contains importance of the presence of the terminal amino acid alanine the last two amino acids, was an effective competitor (spl, sp8, and the no peptide control, —, are shown). (B) The carboxyl terminus is and the terminal carboxylate group as a prerequisite for anti- required for patient antibody binding to rhodopsin. Removing the last body recognition because the synthetic dodecamer sp8, minus amino acid (—1) or amidating the C-terminal carboxylic acid (am) either the terminal amino acid alanine, or with an amidated renders sp8 an ineffective competitor (no peptide control, —). Lower carboxyl-terminus, did not compete for rhodopsin binding. and upper arrows indicate positions of rhodopsin monomer and dimer The C terminus of rhodopsin is not the only, or even the bands. most immunogenic, portion of the rhodopsin molecule, based

Downloaded from iovs.arvojournals.org on 10/01/2021 IOVS, May 1999, Vol. 40, No. 6 Reports 1279

A. B. ruses such as Escberichia coli, human adenovirus 2, and the Pseudomonas strains aemginosa and putida. We add, how- p1 p9 sp1sp7sp8 - am sp8 ever, that we have no direct evidence for previous infection of our patients by these or other organisms that may initiate an aberrant anti-rhodopsin immune response. Although glaucoma is traditionally considered a disease I affecting retinal ganglion cells, a substantial number of electro- physiological studies demonstrate that patients with primary open-angle glaucoma (POAG) may have retinal abnormalities in the outer retina and photoreceptor layer. These studies have mostly identified cone or photopic visual abnormalities in patients with glaucoma.21'~2'f In addition, morphologic exami- nation of eyes with POAG have confirmed selective photore- ceptor injury in these eyes, localized either to specific loss of the red/green cones25 or generalized loss of photoreceptor outer elements.26 However, at least one detailed morphologic study did not find evidence for photoreceptor loss in POAG.27 It is possible that these functional changes occur in the ab- sence of pathomorphology. Although it is unclear how antibodies to the C terminus to rhodopsin may be related to glaucoma pathogenesis, the C terminus is a functionally important part of the rhodopsin I. molecule. Both transducin and arrestin interact with the C- terminal region, and this region is the substrate for rhodopsin FIGURE 4. (A) Additional patient sera contain anti-rhodopsin antibod- kinase.2829 Interruption of any of these protein-protein inter- ies directed against the C-terminal region. In each of three patients, actions would be expected to result in some form of altered only peptides containing the C terminus of rhodopsin (p9, sp8) effec- signaling, Recent studies have indicated that the C terminus tively compete for antibocly-rhodopsin binding. Other peptides tested were ineffective (illustrated for pi, spl, and no peptide control, —). (B) itself is critically important for rhodopsin trafficking through the Golgi apparatus and targeting to the disc membrane of the Additional patient sera contain anti-rhodopsin antibodies directed 30 31 against the C terminus itself. In each of three patients, only the peptide rod photoreceptor. ' Interruption of this process would be containing the C-terminal carboxylic acid (sp8) effectively competes expected to result in some form of cellular pathology. It is for antibocly-rhodopsin binding. Removing the last amino acid (— 1) or intriguing that the heat-shock proteins a- and j3-crystallin, amidating the carboxylic acid (amide) prevents competition. which are expressed in photoreceptors and may participate in the processing and trafficking of rhodopsin,32 also are the target of autoantiboclies in glaucoma patients,10 1516 on studies in sheep, rabbits, and mice. For both sera and In summary, antibodies from all four NPG patients tested monoclonal antibodies to rhodopsin, the majority of antigenic had identical epitope specificity that was localized to the C binding sites were localized in the amino terminal half of the terminus of the rhodopsin molecule. This may signify a com- protein. This may indicate that in our patients the eliciting mon molecular mechanism for anti-rhodopsin antibody gener- immunogen is unlikely to have been intact endogenous rho- ation in these patients. We suggest that these results add to the dopsin. increasing body of evidence that aberrant autoimmunity may The rhodopsin antibodies present in our patients have occur in patients with NPG. Further studies of the molecular remained detectable at approximately the same titers from the nature of the autoantigens and autoantiboclies in similar pa- time of their earliest detection to the present, with a follow-up tients may contribute to our understanding, diagnosis, and period of at least 5 years for the patients reported here. A treatment of glaucoma. potential explanation for the immunoreactivity observed to rhodopsin in patients with NPG may be molecular mimicry. A premise of molecular mimicry is that immune responses to References infectious agents may generalize to native cellular proteins 1. Sommer A. Intraocular pressure and glaucoma. AniJ Opbtbahnol. with similar epitope homology, resulting in serum antibodies 1989;107:186-188. that recognize these proteins. The hypothesis whereby infec- 2. Sommer A. Doyne Lecture. Glaucoma: facts and fancies. Eye. tious agents initiate aberrant autoimmune tissue-specific re- 1996;10:295-301. sponses due to molecular mimicry between the infectious 3. Kerrigan LA, Zack DJ, Quigley HA, Smith SD, Pease ME. TUNEL- agent and native tissue proteins is well recognized1718 and has positive ganglion cells in human primary open-angle glaucoma. been implicated in the development of numerous organ-spe- Arch Ophthalmol. 1997;115:1031-1035. 19 4. Quigley HA, Nickells RW, Kerrigan LA, et al. Retinal ganglion cell cific autoimmune diseases. If mimicry were operative in death in experimental glaucoma and after axotomy occurs by some patients with NPG, antibodies made in response to ex- apoptosis. Invest Ophthalmol Vis Sci. 1995;36:774-786. posure to certain infectious agents might cross-react with a 5. Garcia-Valenzuela E, Shareef S, Walsh J, Sharma SC. Programmed protein such as rhodopsin, an abundant immunoreactive reti- cell death of retinal ganglion cells during experimental glaucoma. nal protein. Of particular interest is the finding that the C Exp Eye Res. 1995;6l:33-44. terminus of rhodopsin shares sequence identity to the C ter- 6. Wax MB, Tezel G, Edward FD. Clinical and ocular histopathologi- cal findings in a patient with normal-pressure glaucoma. Arch mini of numerous proteins from pathogenic bacteria and vi- Ophthalmol. 1998;l 16:993-1001.

Downloaded from iovs.arvojournals.org on 10/01/2021 1280 Reports IOVS, May 1999, Vol. 40, No. 6

7. Cartwright MJ, Grajewski AL, Friedberg ML, Anderson DR, Rich- 22. Fazio DT, Heckenlively JR, Martin DA, Christensen RE. The elec- ards DW. Immune-related disease and normal-tension glaucoma. troretinogram in advanced open-angle glaucoma. Doc Ophthal- Arch Ophthalmol. 1992; 110:500-502. mol, 1986;63:45-54. 8. Wax MB, Barrett DA, Pestronk A. Increased incidence of parapro- 23. Odom JV, Feghali JG, JinJC, Weinstein GW. Visual function deficits teinemia and autoantibodies in patients with normal pressure in glaucoma. Electroretinogram pattern and luminance nonlineari- glaucoma. Am JOphthalmol. 1994;117:56l-578. ties. Arch Ophthalmol. 1990;108:222-227. 9. Wax MB, Tezel G, Saito I, et al. Anti-Ro/SS-A positivity and heat 24. Holopigian K, Seiple W, Mayron C, Koty R, Lorenzo M. Electro- shock protein antibodies in patients with normal-pressure glau- physiological and psychophysical flicker sensitivity in patients coma. Am J Ophthalmol. 1998;125:l45-157. with primary open-angle glaucoma and ocular hypertension. In- 10. Tezel G, Seigel G, Wax MB. Autoantibodies to small heat shock vest Ophthalmol Vis Sci. 1990:31:1863-1868. proteins in glaucoma. Invest Ophthalmol Vis Sci. 1998;39:2277- 25- Nork TM, Poulsen GL, Vaegen, Sarks S. Photoreceptor damage and 2287. loss in human eyes with primary open angle glaucoma [ARVO 11. Romano C, Barrett DA, Li Z, Pestronk A, Wax MB. Anti-rhodopsin Abstract]. Invest Ophthalmol Vis Sci. 1995;36(4):S331. Abstract nr antibodies in sera from patients with normal-pressure glaucoma. 1546. Invest Ophthalmol Vis Sci. 1995; 36:1968 -1975. 26. Janssen P, Naskar R, Moore S, Thanos S, Thiel HJ. Evidence for 12. Tezel G, Kass MA, Kolker AE, Wax MB. Comparative optic disc glaucoma-induced horizontal cell alterations in the human retina. analysis in normal pressure glaucoma, primary open-angle glaucoma, GerJ Ophthalmol. 1996;5:378-385. and ocular hypertensioa Ophthalmology. 1996;103:2105-2113. 27. Kendell KR, Quigley HA, Kerrigan LA, Pease ME, Quigley EN. 13. Adamus G, Zam ZS, Arendt A, et al. Anti-rhodopsin monoclonal antibodies of defined specificity: characterization and application. Primary open-angle glaucoma is not associated with photoreceptor Vision Res. 1991;31:17-31. loss. Invest Ophthalmol Vis Sci. 1995:36:200-205. 14. Litman BJ. Purification of rhodopsin by concanavalin A affinity 28. Hargrave PA, McDowell JH. Rhodopsin and phototransduction: a chromatography. MethodsEnzymol. 1982;81:150-153. model system for G protein-linked receptors. FASEB J. 1992;6: : 15- Hargrave PA, Adamus G, Arendt A, et al. Rhodopsin's amino ter- 2323 2331. minus is a principal antigenic site. Exp Eye Res. 1986;42:363-373- 29- Hurley JB, Spencer M, Niemi GA. Rhodopsin phosphorylation and 16. Adamus G, Arendt A, Hargrave PA. Genetic control of antibody its role in photoreceptor function. Vision Res. 1998:38:1341- response to bovine rhodopsin in mice: epitope mapping of rho- 1352. dopsin structure. / Neuroimmunol. 1991;34:89-97. 30. Deretic D, Schmerl S, Hargrave PA, Arendt A, McDowell JH. Reg- 17. Marrack P, Kappler J. Subversion of the immune system by patho- ulation of sorting and post-golgi trafficking of rhodopsin by its C- gens. Cell. 1994;76:323-332. terminal sequence QVS(A)PA. Proc Nat/ Acad Sci USA. 1998;95: 18. Oldstone MB. Molecular mimicry and autoimmune disease. Cell. 10620-10625. 1987;50:819-820. 31. Sung CH, Makino C, Baylor D, Nathans J. A rhodopsin gene muta- 19. Young DB. Heat-shock proteins: immunity and autoimmunity. tion responsible for autosomal dominant retinitis pigmentosa re- Curr Opin Immunol. 1992;4:396-400. sults in a protein that is defective in localization to the photore- 20. Alvis DL. Electroretinographic changes in controlled chronic open- ceptor outer segment./Neurosci. 1994;l4:5818-5833. angle glaucoma. AmJ Ophthalmol. 1966;61:121-131. 32. Deretic D, Aebersold RH, Morrison HD, Papermaster DS. Alpha A- 21. Bartl G. The electroretinogram and the visual evoked potential in and alpha B-crystallin in the retina: association with the post-Golgi normal and glaucomatous eyes. Albrecht Von Graefes Arch Klin compartment of frog retinal photoreceptors. / Biol Chem. 1994; Exp Ophthalmol. 1978;207:243-269. 269:16853-16861.

Mice Deficient in Inducible Nitric required to effect the pathogenesis of experimental auto- immune uveoretinitis (EAU) in mice. Oxide Synthase Are Susceptible METHODS. Mice with a homologous disruption of the iNOS to Experimental Autoimmune gene (iNOS KO) were evaluated for their ability to develop EAU and associated cellular responses after immunization Uveoretinitis with the interphotoreceptor retinoid-binding protein. EAU was determined by histopathology 21 days after Phyllis B. Silver, Teresa K. Tarrant, uveitogenic immunization, and antigen-specific cellular Chi-Chao Chan, Barbara Wiggert, and responses were assessed by delayed type hypersensitivity Rachel R. Caspi and lymphocyte proliferation. RESULTS. iNOS knockout (iNOS KO) mice developed EAU PURPOSE. Nitric oxide (NO) is an important mediator of with scores similar to wild-type mice and exhibited good inflammatory tissue damage. The present study addresses cellular responses to the immunizing antigen. the question whether inducible nitric oxide synthase CONCLUSIONS. A functional iNOS gene is not necessary for (iNOS), and consequently the ability to upregulate NO, is EAU pathogenesis. Therefore, upregulation of NO is not required to mediate autoimmune tissue damage in the eye. {Invest Ophthalmol Vis Sci. 1999;40:1280-1284) From the Laboratory of Immunology, National Eye Institute, Na- tional Institutes of Health, Bethesda, Maryland. Submitted for publication May 27, 1998; revised January 5, 1999; xperimental autoimmune uveoretinitis (EAU) is a T-cell- accepted January 20, 1999. mediated autoimmune disease model that targets the neu- Proprietary interest category: N. E Reprint requests: Rachel R. Caspi, Laboratory of Immunology, ral retina. EAU can be induced in mice, rats, and primates by National Eye Institute, NIH, Building 10, Room 10N222, 10 Center Dr immunization with retinal antigens or their fragments, or by MSC 1857, Bethesda, MD 20892. the adoptive transfer of retinal antigen-specific Thl-like

Downloaded from iovs.arvojournals.org on 10/01/2021