MECHANISMS OF OPHTHALMIC DISEASE

SECTION EDITOR: LEONARD A. LEVIN, MD, PHD Myocilin and A TIGR by the Tail?

Douglas H. Johnson, MD

n 1997, Stone and 14 colleagues from 7 laboratories reported the identification of a (TIGR) associated with juvenile open-angle glaucoma (JOAG). Screening of adults with pri- mary open-angle glaucoma (POAG) revealed that about 4% also carried a mutation of the coding region of this gene. The mutations were found through genetic linkage analysis of Ifamilies with JOAG. Juvenile open-angle glaucoma was a logical starting point in the search for genetic causes of open-angle glaucoma: it shows a strong autosomal-dominant inheritance pat- tern, occurs at an early age, demonstrates obvious phenotypic signs (dramatic elevation of intra- ocular pressure and subsequent optic nerve damage), and is likely to be found in multiple genera- tions as parents of affected children are still living. These factors, however, also serve to distinguish it from adult-onset POAG, which generally has a lower intraocular pressure and a less severe course. The discovery of the actual gene represented a true advance over previous studies that had mapped the gene to a segment of a but did not identify the specific gene. How the mutant gene causes glaucoma is unknown and is the subject of intense research. To date, 26 mutations in the TIGR gene sequence (the term TIGR has been replaced by the term myocilin, abbreviated MYOC) have been described, all associated with either JOAG or adult-onset POAG. A correlation between specific mutations in MYOC and the clinical course of glaucoma has been found. Not all cases of JOAG or POAG have mutations in the MYOC gene, however, indicating that more discoveries of other are yet to come. Arch Ophthalmol. 2000;118:974-978 A new door has opened in glaucoma re- netic linkage studies.2 Glaucoma is indi- search. The discovery of the myocilin cated by GLC, with “1” indicating primary gene and its association with glaucoma open-angle glaucoma (POAG) and “A” in- provides a key to understanding the physi- dicating that it was the first gene linkage site ology of aqueous outflow through the found in POAG. As expected, other gene trabecular meshwork, and the patho- linkages for glaucoma have been identi- physiology of glaucoma. Undoubtedly fied: 5 additional POAG linkages (GLC1B, the answers will be far more complex than GLC1C, GLC1D, GLC1E, and GLC1F), con- we can imagine at our current state of genital glaucoma (GLC3A and GLC3B), and knowledge. developmental anomalies or other condi- tions associated with glaucoma (Rieger syn- WHAT’S IN A NAME? drome, RIEG 1; pigment dispersion syn- drome, PDS).3 Some of the candidate genes GLC1A are transcription factors that affect differ- ent stages of embryonic development.3 The region Stone et al1 identified was first given the nonspecific name GLC1A, follow- TIGR ing the convention for naming genes not yet discovered but surmised because of ge- The mutant gene at the GLC1A locus iden- tified by Stone and collegues for juvenile From the Mayo Clinic, Rochester, Minn. open-angle glaucoma (JOAG) was found to

(REPRINTED) ARCH OPHTHALMOL / VOL 118, JULY 2000 WWW.ARCHOPHTHALMOL.COM 974

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 be a previously described gene, TIGR (trabecular mesh- Olfactomedin work–induced glucocorticoid response). This gene had Domain Domain 1 50 208 328 502 504 been found, sequenced, and named during studies on cor- NH COOH ticosteroid-induced glaucoma. Polansky et al4,5 had found 2 Mutation a series of that were induced by long-term dexa- Zipper } Sites methasone treatment in cultured human trabecular cells. One of the proteins, originally called TIGR, had the in- Figure 1. Schematic diagram of human myocilin . Note that the myosin-like domain is near the N terminal, and the olfactomedin-like domain triguing feature of time-dependent induction: progres- is near the C terminal. Reported mutation sites are indicated by the short sively more protein was produced with continued dexa- lines at the bottom of the protein. Most mutations are in the methasone treatment. This time course of induction was olfactomedin-like domain. The amino group is NH2; carboxyl group, COOH. similar to that of the clinical situation of corticosteroid- induced glaucoma, occurring over a period of several weeks. glaucomatous eyes, it may be present in more regions and Subsequent analysis through a complementary DNA li- have more intense labeling in glaucomatous eyes.19 The brary screening enabled discovery and cloning of the gene.4-6 presence of MYOC in normal eyes is consistent with a physiologic role for MYOC and indicates that the mere Myocilin presence of MYOC does not by itself cause glaucoma.

While the study on JOAG and TIGR by Stone et al1 was in PROPERTIES OF MYOC PROTEIN press, Kubota et al7 in Japan submitted a study on the dis- covery of a protein associated with the in the Myocilin is a glycoprotein that exists in glycosylated and retina. They termed this protein myocilin (MYOC) be- nonglycosylated forms, with molecular weights of 66 kd cause it shared homologous regions with myosin.7 This and 55 kd.4-6 It is present both intracellularly and extra- was the same gene and protein that Nguyen and col- cellularly.4-6 Myocilin protein consists of 504 amino ac- leagues6 had described and named TIGR, unbeknownst to ids and has 2 major domains: a myosin-like domain near Kubota. In 1998 the Organization Ge- the N terminal, and an olfactomedin-like domain near nome Database Nomenclature Committee assigned the the C terminal (Figure 1).4-6 Most of the mutations in gene the name myocilin, and it is now referred to as MYOC MYOC associated with glaucoma are in the olfactome- rather than TIGR. Note the convention of using italics to din domain.1,2 Olfactomedins are a family of mucus pro- refer to the gene and standard font to refer to the protein. teins that are predominately found in nasal mucus.20 Myo- cilin protein has several noteworthy elements, including AN ENDOGENOUS PROTEIN THAT (1) a “” region: this is a series of 7 leucine INCREASES DURING STRESS molecules (amino acids) in a configuration that might allow it to interact with leucine zippers present in other Myocilin is normally present in a variety of ocular and proteins; using these zippers, MYOC could also form nonocular tissues: trabecular meshwork, cornea, retina, dimers or oligomers with itself; (2) multiple possible sites optic nerve, ciliary nerves, and the heart, skeletal muscle, for and phosphorylation; (3) presumed bind- stomach, thyroid, bone marrow, prostate, intestine, lung, ing sites for hyaluronic acid and heparan sulfate4,7,21; and pancreas, and lymph node.8,9 In the eye, MYOC may be (4) a “signal sequence” of 32 amino acids at the N ter- produced in greater amounts in times of stress. It is pres- minal of the molecule, which is usually found in mol- ent in increased amounts in cultured anterior segments ecules secreted extracellularly. from human eyes and in monolayer cultured trabecular The structure of MYOC protein is conserved through cells after undergoing dexamethasone treatment, oxida- evolution, with 80% or more homogeneity of amino ac- tive stress, stretching, and treatment with transforming ids between human and mouse MYOC.8,9 This cross- growth factor ␤.4-6,10-12 As such, MYOC is one of a di- species presence is in keeping with a presumed physi- verse group of stress proteins, the role of which seems ologic role for MYOC, though the role is unknown. to be protective of vital cellular proteins or enzymes.13-16 Comparison of the sequence of MYOC across species can Another class of stress-induced proteins that have re- be helpful in understanding which domains are of func- ceived study are the heat shock proteins, first found in tional importance. The preservation of the leucine zip- the larva of the fruit fly Drosophila when enlargements per motifs and sites for glycosylation and phosphoryla- of certain regions of the were noted after tion between mouse and man suggests that these play a heating (hence the term heat shock proteins).17 Some of vital role in the function of the protein, while regions not the stress proteins are thought to bind to important cel- conserved across species are presumably not as critical lular proteins, preventing them from unfolding or dena- to its function.9 turing under stress, earning them the name molecular The normal physiologic function of MYOC in the chaperones.13-16 Such stress proteins may be endog- cell is unknown. It may serve a structural function within enously present in a variety of tissues and also may have the cytoplasm, or it may associate with other molecules specific functions of their own: ␣-B crystalline is a ma- within the cell, perhaps as a molecular chaperone. Ex- jor component of the crystalline lens and is also present tracellularly, it may be involved in creating resistance to in smaller amounts in other tissue (cardiac, lung, kid- aqueous outflow by binding to other extracellular mol- ney, muscle, brain, and retina), presumably in a protec- ecules or to the cell membrane of trabecular cells. tive role as a molecular chaperone.13-16,18 Although MYOC In addition to studies of the MYOC protein and gene, is present in the trabecular meshwork of both normal and attention has focused on the promoter region of MYOC.

(REPRINTED) ARCH OPHTHALMOL / VOL 118, JULY 2000 WWW.ARCHOPHTHALMOL.COM 975

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 Figure 2. Juvenile open-angle glaucoma, trabecular meshwork. Fingerprint material resembling basement membrane is present in the outer comeoscleral region (large arrows). Small arrow points to trabecular cell (original magnification: left, ϫ3900; right, ϫ26000). Reprinted with permission from Ophthalmologica (1997;211:140-146). Copyright 1997, S. Karger AG.25

SC SC

SD SD

3 µm 3 µm

Figure 3. Corticosteroid glaucoma, trabecular meshwork. Fingerprint Figure 4. Primary open-angle glaucoma. Accumulation of tendon and material resembling basement membrane (FBM material) is present in the tendon-sheath material (SD plaques) under the inner wall of Schlemm’s inner juxtacanalicular region (arrows). Inset shows higher power view of canal (SC). This specimen contains unusually dense and heavy FBM material. SC indicates Schlemm’s canal (original magnification ϫ4000; accumulations of this material. Contrast this appearance with that of Figure 2 inset, ϫ17000). and Figure 3.

Promoters are gene elements “upstream” from the protein- ent appearance than that of eyes with POAG when ex- encoding DNA sequence. They control the amount of pro- amined with the electron microscope. Histologic stud- tein produced by the gene by regulating transcription of ies of JOAG report the accumulation of an abnormal the protein-coding portions of the DNA. The promoter basement membrane-like material within the trabecular region of MYOC has been identified and contains re- meshwork, which may fill the meshwork and disrupt gions (consensus motifs) that are responsive to gluco- Schlemm’s canal.23-25 This differs dramatically from the corticoids and a number of other hormones.4-6 The pres- changes of adult-onset POAG: thickening of tendon and ence of this glucocorticoid response element may explain tendon-sheath material within the meshwork.26,27 The the increase in the amount of normal MYOC protein pres- changes of JOAG are strikingly similar to those of cor- ent in cultured trabecular cells from susceptible pa- ticosteroid-induced glaucoma in which the meshwork also tients after corticosteroid treatment.4-6,22 contains accumulations of abnormal basement membrane- like material, often in whorls resembling a fingerprint JOAG, POAG, MYOC, AND STEROIDS (Figures 2, 3, and 4).28,29 Corticosteroid treatment causes the overproduc- Juvenile open-angle glaucoma has an earlier age of on- tion of MYOC in cultured trabecular cells from some pa- set, higher intraocular pressure (IOP), and more severe tients.4-6 Could excess MYOC be a link between steroid course than adult-onset POAG. In addition, trabecular glaucoma and JOAG? Does the similarity in the ultra- meshwork from eyes with JOAG has a markedly differ- structural appearance of the 2 conditions suggest that

(REPRINTED) ARCH OPHTHALMOL / VOL 118, JULY 2000 WWW.ARCHOPHTHALMOL.COM 976

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 the abnormal basement membrane–like material con- chromosome carries the mutant gene, and presumably sists, at least in part, of MYOC? Myocilin is known to be 1 mutant MYOC molecule and 1 normal MYOC mol- a sticky molecule with binding sites for several compo- ecule are produced. The mutant MYOC protein may have nents of the basement membrane.4-6 a different shape (secondary or tertiary protein struc- ture) that prevents its interaction with the normal MYOC HOW COULD MYOC CAUSE GLAUCOMA? molecule. Morrissette and colleagues have hypoth- esized that if both copies of MYOC are mutant (homo- MYOC Alone zygous), the MYOC proteins may interact and no glau- coma would result.30 If excess MYOC accumulated in the trabecular mesh- work, whether from overproduction or a decrease in deg- MYOC Plus radation, glaucoma could develop. As discussed earlier, corticosteroid treatment causes the overproduction of nor- None of these speculations account for the relatively de- mal MYOC protein in susceptible trabecular meshwork layed onset of glaucoma even in the juvenile-onset cases: cells. This MYOC protein is found both intracellularly Why is glaucoma not present at birth or in the first few and extracellularly.4-6 Excess extracellular MYOC could years of life, given that the mutation is present at birth? bind to the aqueous outflow pathways and increase out- In addition, not all patients with mutant MYOC have glau- flow resistance, similar to the mechanism of retained hy- coma. At least 10 cases of adult patients with the aluronic acid (Healon; Pharmacia & Upjohn Inc, Kalama- Gln368STOP mutation who do not have glaucoma have zoo, Mich) or other viscoelastic material increasing IOP been reported.2,31 Both of these lines of evidence suggest after cataract surgery. Preliminary results with recombi- that a second factor may be involved in causing glau- nant MYOC protein in my laboratory, in conjunction with coma or that there may be a cascade of events leading to the studies of Nguyen et al6 and Polansky et al,4,5 con- glaucoma. This second factor could be genetic or envi- firm that recombinant MYOC protein can indeed el- ronmental: it could be an abnormality in the promoter evate IOP in eyes in perfusion organ culture of the an- region, an abnormality in other molecules with which terior segment. If synthesis of a mutant MYOC protein MYOC presumably interacts, changes in the outflow path- escapes control of normal feedback regulation, then ex- way by light-induced oxidation, or the breakdown of a cess amounts of MYOC or its mutant form could be pro- redundant pathway that had been compensating for the duced and thus increase outflow resistance. physiologic function of mutant MYOC protein. Alternately, the mutant form of MYOC may be un- Despite the unanswered questions, the discovery of able to fulfill the presumed physiologic role of MYOC the MYOC gene and its link to open-angle glaucoma rep- protein. In this scenario, some mutations would cause resent a true step forward in the search for the patho- more problems than others. This is supported by the find- genesis of glaucoma. Study of MYOC and its function in ing of a clinical correlation among the various muta- health and disease will lead to a better understanding of tions and the onset and severity of glaucoma. The most the mechanism of aqueous outflow and ultimately to common mutation, Gln368STOP, is associated with an therapy that could correct the genetic problems associ- older age of onset and less elevation of IOP than the re- ated with glaucoma. ported missense mutations described in the younger- onset groups.2 This correlation between genotype and phe- Accepted for publication March 27, 2000. notype and the delayed onset of glaucoma suggest that a Corresponding author: Douglas H. Johnson, MD, Mayo truncated form of the MYOC protein may still serve some Clinic, 200 First St SW, Rochester, MN 55905 (e-mail: physiologic function, while those mutant forms with varia- [email protected]). tions in the sequence are more problematic and cause glaucoma at an earlier age. One could imag- REFERENCES ine that proper dimerization or oligomerization is nec-

essary for the function of MYOC and that certain muta- 1. Stone EM, Fingert JH, Alward WLM, et al. Identification of a gene that causes tions prevent proper interactions of this protein. The primary open-angle glaucoma. Science. 1997;275:668-670. resultant increase in monomeric forms of the protein could 2. Alward WLM, Fingert JH, Coote MA. Clinical features associated with mutations be associated with increased resistance in the trabecular in the open-angle glaucoma gene (GLCIA). N Engl J Med. 1998; meshwork. Not all changes in the MYOC gene sequence 338:1022-1027. 3. Sarfarazi M. Recent advances in molecular genetics of . Hum Mol Genet. have been associated with glaucoma; those apparently 1997;6:1667-1677. nondisease-causing sequence variations are termed poly- 4. Polansky JR. HTM cell culture model for steroid effects on IOP: overview. Lu¨tjen- morphisms. Drecoll E, ed. Basic Aspects of Glaucoma Research III. Stuttgart, Germany: Schat- Another interesting possibility raised by the find- tauer; 1993:307-318. ing of Morissette et al30 is that patients with the same 5. Polansky JR, Fauss DJ, Chen P, et al. Cellular pharmacology and molecular bi- ology of the trabecular meshwork inducible glucocorticoid response gene prod- MYOC mutation in both chromosomes (maternal and pa- uct. Ophthalmologica. 1997;211:126-139. ternal) do not develop glaucoma while patients with a 6. Nguyen TD, Chen P, Huang WD, Chen H, Johnson D, Polansky JR. Gene struc- mutation in only 1 of 2 chromosomes will develop glau- ture and properties of MYOC, an olfactomedin-related glycoprotein cloned from coma. This implies that the interaction between the MYOC glucocorticoid-induced trabecular meshwork cells. J Biol Chem. 1998;273:6341- 6350. protein molecules produced by each chromosome may 7. Kubota R, Noda S, Wang Y, et al. A novel myosin-like protein (myocilin) ex- be important to the function of MYOC. In the usual case pressed in the connecting cilium of the photoreceptor: molecular cloning, tissue of glaucoma associated with a MYOC mutation, only 1 expression, and chromosomal mapping. Genomics. 1997;41:360-369.

(REPRINTED) ARCH OPHTHALMOL / VOL 118, JULY 2000 WWW.ARCHOPHTHALMOL.COM 977

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 8. Tomarev SI, Tamm ER, Chang B. Characterization of the mouse Myoc-MYOC gene. 21. Ortego J, Escribano J, Coca-Prados M. Cloning and characterization of sub- Biochem Biophys Res Commun. 1998;245:887-893. tracted cDNAs from a human ciliary body library encoding MYOC, a protein in- 9. Fingert JH, Ying L, Swiderski RE, et al. Characterization and comparison of the volved in juvenile open-angle glaucoma with homology to myosin and olfacto- human and mouse GLC1A glaucoma genes. Genome Res. 1998;8:377-384. medin. FEBS Lett. 1997;413:349-353. 10. Tamm ER, Tomarev SI, Russell P, Epstein DI, Piatigorsky J. MYOC expression 22. Chen H, Kitazawa Y, Kawase K, et al. Identification of a promoter sequence vari- and its transcriptional control in human and mouse trabecular meshwork [ab- ant in steroid responders and evidence for a glucocorticoid-induced DNA bind- stract]. Invest Ophthalmol Vis Sci. 1998;39:S34. ing protein of the MYOC gene in TM cells [abstract]. Invest Ophthalmol Vis Sci. 11. Tamm ER, Tomarev SI, Russell P, Epstein DL, Piatigorsky J. MYOC/MYOC ex- 1999;40:S505. pression in human and mouse trabecular meshwork. Exp Eye Res. 1998;67 23. Tawara A, Inomata H. Developmental immaturity of the trabecular meshwork in (suppl):S135. juvenile glaucoma. Am J Ophthalmol. 1984;98:82-97. 12. Tamm ER, Russell P, Epstein DL, Johnson D, Piatigorsky J. Modulation of myocilin/ 24. Tawara A, Inomata H. Distribution and characterization of sulfated proteogly- MYOC expression in human trabecular meshwork. Invest Ophthalmol Vis Sci. cans in the trabecular tissue of goniodysgenetic glaucoma. Am J Ophthalmol. 1999;40:2577-2582. 1994;117:741-755. 13. Ingolia TD, Craig EA. Four small Drosophila heat shock proteins are related to 25. Furuyoshi N, Furuyoshi M, Futa R, Gottanka J, Lu¨tjen-Drecoll E. Ultrastructural each other and to mammalian ␣-crystallin. Proc Natl Acad SciUSA. 1982;79: changes in the trabecular meshwork of juvenile glaucoma. Ophthalmologica. 1997; 2360-2364. 211:140-146. 14. Klemenz R, Fro¨hli E, Steiger RH, Scha¨fer R, Aoyama A. ␣-B-crystallin is a small 26. Rohen JW. Why is intraocular pressure elevated in chronic simple glaucoma? heat shock protein. Proc Natl Acad SciUSA. 1991;88:3652-3656. anatomical considerations. Ophthalmology. 1983;90:758-765. 15. Ellis J, van der Vies SM. Molecular chaperones. Ann Rev Biochem. 1991;60: 27. Gottanka J, Johnson DH, Martus P, Lu¨tjen-Drecoll E. Severity of optic nerve dam- 321-347. age in eyes with POAG is correlated with changes in the trabecular meshwork. 16. Wang K, Spector A. ␣-Crystallin can act as a chaperone under conditions of oxi- J Glaucoma. 1997;6:123-132. dative stress. Invest Ophthalmol Vis Sci. 1995;36:311-321. 28. Rohen JW, Linner E, Witmer R. Electron microscopic studies on the trabecular 17. Ritossa F. A new puffing pattern induced by temperature shock and DNP in Dro- meshwork in two cases of corticosteroid glaucoma. Exp Eye Res. 1973;17:19- sophila. Experientia. 1962;18:571-573. 31. 18. Pineda II R, Chan C-C, Ni M, et al. Human retinoblastoma cells express ␣-B- 29. Johnson D, Gottanka J, Flu¨gel C, Hoffmann F, Futa R, Lu¨tjen-Drecoll E. Ultra- crystallin in vivo and in vitro. Curr Eye Res. 1993;12:239-245. structural changes in the trabecular meshwork of human eyes treated with cor- 19. Lu¨tjen-Drecoll E, May CA, Polansky JR, Johnson DH, Bloemendal H, Nguyen TD. ticosteroids. Arch Ophthalmol. 1997;115:375-383. Localization of the stress proteins ␣-B-crystallin and trabecular meshwork in- 30. Morissette J, Clepet C, Moisan S, et al. Homozygotes carrying an autosomal ducible glucocorticoid response protein in normal and glaucomatous trabecular dominant MYOC mutation do not manifest glaucoma. Nat Genet. 1998;19: meshwork. Invest Ophthalmol Vis Sci. 1998;39:517-525. 319-321. 20. Snyder AD, Rivers AM, Yokoe H, Menco BPM, Anholt RH. Olfactomedin: purifi- 31. Mardin CY, Michels-Rautenstrauss K, O¨ zbey S, Rautenstrauss B, Wisse M, Nau- cation, characterization, and localization of a novel olfactory glycoprotein. Bio- mann GOH. GLN368STOP mutation in the GLCIA-gene of a German family [ab- chemistry. 1991;30:9143-9153. stract]. Invest Ophthalmol Vis Sci. 1999;40:S78.

From the Archives of the ARCHIVES

A look at the past...

yclodialysis as an operation for the relief of increased intraocular pres- sure has gained in favor with the staff of the ophthalmologic clinic at Cthe University of Iowa until at present it is our operation of choice for chronic noncongestive wide angle or narrow angle glaucoma, hydroph- thalmos and certain types of secondary glaucoma, notably that after cataract extraction.

Reference: O’Brien CS, Weih J. Cyclodialysis. Arch Ophthalmol. 1949;42:606.

(REPRINTED) ARCH OPHTHALMOL / VOL 118, JULY 2000 WWW.ARCHOPHTHALMOL.COM 978

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021