Cornea 19(5): 699–711, 2000. © 2000 Lippincott Williams & Wilkins, Inc., Philadelphia Genetics of the Corneal Dystrophies What We Have Learned in the Past Twenty-five Years Anthony J. Bron, F.R.C.S. Purpose. To indicate important changes in our understanding of The term corneal dystrophy will be used here to define primary, the corneal dystrophies. Methods. A review of the literature of the inherited, bilateral disorders of the cornea affecting transparency last quarter of a century. Results. The earliest clinical classifica- or refraction. Such disorders may or may not disturb vision. At tions of the corneal dystrophies were based on the application of various times in the past they have been said to be of early onset, clinical, biological, histochemical, and ultrastructural methods. axially symmetrical, slowly progressive, or stationary. Earlier defi- Since then, the first great impetus to our understanding has come nitions also required that the affected corneas be free from vascu- from the application of techniques to map disorders to specific chromosome loci, using polymorphic markers. More recently, us- larization and not associated with systemic disease. These are rea- ing candidate gene and related approaches, it has been possible to sonable guidelines, which draw attention to the possibility of a identify genes causing several of the corneal dystrophies and the corneal dystrophy in a given clinical situation. But there are ex- mutations responsible for their phenotypic variation. A notable ceptions to each of these characteristics in conditions generally success has been to show that several important “stromal” dystro- accepted to be corneal dystrophies. Thus, asymmetrical or even phies result from mutations in the gene ␤ig-h3, which encodes for unilateral Meesmann’s and lattice dystrophy are reported, and vas- the protein keratoepithelin (␤ig-h3). Conclusions. For the corneal cularization is a characteristic of autosomal dominant keratitis and, dystrophies, as with other inherited disorders, there is room for two eventually, a feature of gelatinous drop-like dystrophy (GDLD). sorts of classification system, one based mainly on clinical pre- Haugen and Bertelsen3 have described a condition in a mother and sentation and the other on an up-to-date understanding of the ge- two sons in whom a dysplastic fibrovascular condition of the cor- netic mechanisms. They are not mutually exclusive. Some devel- opmental corneal disorders are also discussed. neas is accompanied by keloidal deformities of the hands. Cornea Key Words: Corneal dystrophies—Genetics—Developmental guttata and Fuchs’ dystrophy are late in presentation, and lattice disorders. corneal dystrophy (LCD) type II is a manifestation of an inherited, systemic, amyloid polyneuropathy. Macular corneal dystrophy (MCD) is accompanied by changes in levels of antigenic serum Our perceptions of the corneal dystrophies have changed, dra- keratan sulfate, implying extraocular involvement. Other corneal matically over the past quarter of a century, as new genetic infor- disorders, too, are recognized features of systemic diseases, such mation has become available. Over 10 chromosome loci have been as X-linked ichthyosis (steroid sulfatase deficiency),4 ceramide linked to specific corneal dystrophies, and mutations in four genes trihexosidase deficiency (Fabry disease),5 the mucopolysacchari- are known to be responsible for some of these. Multiple mutations doses, and the mucolipidoses. in the gene ␤ig-h3, which encodes for the protein keratoepithelin (␤ig-h3), have been invoked to explain the phenotypic diversity of the classical stromal dystrophies and are providing insights into KERATOEPITHELIN DYSTROPHIES their mechanisms. With these new findings and the expectation of further developments in this field, there has been a call for a Considerable interest has focused on mutations in the kerato- revised classification of the corneal dystrophies, substituting a ge- epithelin gene (␤ig-h3) as the basis for several, phenotypically notypic for a phenotypic approach. Some valuable genotypic clas- distinct corneal dystrophies. Keratoepithelin is an adhesion pro- 1,2 sifications have been published recently. However, the two ap- tein, which is strongly expressed by the corneal epithelium and, to proaches are not mutually exclusive and the phenotypic approach, a lesser extent, by other corneal cells. It is secreted by the epithe- based on an assessment of clinical and structural features, is likely lium and is present in normal stroma, covalently bound to type VI to continue to provide the most comprehensive clinical account of collagen. A study by Streeten et al.6 showed the strongest immu- these disorders for some time to come. The genotypic approach noreactivity for keratoepithelin in Bowman’s layer, at an interla- will provide increasing understanding and, in the long term, the mellar location (to a lesser extent), and at the interface between opportunity for medical intervention, using conventional pharma- stromal collagen and Descemet’s layer (to a minor degree). cologic approaches, or gene therapy. The keratoepithelin molecule is about the size of albumin and, therefore, is sufficiently small to diffuse across the stroma from its Submitted March 7, 2000. Accepted April 30, 2000. epithelial source. Mutated keratoepithelins are of similar size and From Nuffield Laboratory of Ophthalmology, University of Oxford, their diffusability could explain the slow anterior-to-posterior evo- Oxford, U.K. Address correspondence and reprint requests to Dr. Anthony J. Bron, lution of granular, lattice, and Avellino corneal dystrophies over Nuffield Laboratory of Ophthalmology, University of Oxford, Walton time. Although it is not excluded that keratocytes and endothelial Street, Oxford OX2 6A W, U.K. cells may contribute to the formation of deposits in the keratoepi- 699 700 A.J. BRON thelin dystrophies, one of the strongest pieces of evidence support- and where they aggregate or co-aggregate to form amyloid or other ing a major epithelial origin is the initial confinement of recur- deposits. In the case of Thiel-Behnke dystrophy, which is confined rences to the epithelium, after keratoplasty or phototherapeutic to Bowman’s layer, it must be assumed that the aggregates, or the keratoplasty (PTK).7 This has been confirmed ultrastructurally.8 mutated keratoepithelin itself, either have a high affinity for Bow- With time, recurrences spread to affect the stroma.9 man’s layer or that their diffusion is limited by size constraints. It Between 1994 and 1996, four reports appeared, linking granular, may be relevant that immunolocalization of native keratoepithelin lattice,10 granular-lattice (or Avellino dystrophy),11 and Reis- is strongest at the level of Bowman’s layer in the normal cornea.6 Bu¨cklers’ dystrophy,12 to chromosome 5q31.13 It was mooted that The keratoepithelin dystrophies have in common a clear zone of these disorders, hitherto regarded as discrete clinical entities, unaffected, peripheral corneal stroma about 1–2 mm wide. The might be allelic forms. Korvatska et al.14 refined the locus of an basis for this is not known but may reflect removal of the diffusing, assumed allelic gene to a 1CM region of 5q (5q22–5q32) and mutated keratoepithelin by the limbal blood vessels or, alterna- excluded SPARC and LOX as relevant candidate genes. Later, the tively, that epithelial keratoepithelin expression is lower in the same group reported missense mutations in BIGH3, in family periphery, where the mitotic rate is high, than in the central cornea. members of each of the dystrophies mapping to 5q.15 Mutations at This deserves further study. codon 124 were associated with LCD (arginine to cysteine) and Table 1 offers a phenotypic classification of the corneal dystro- Avellino dystrophy (arginine to histidine), whereas mutations at phies supplemented with information about linkage, or genetic codon 555 were associated with granular corneal dystrophy (GCD) information where it is available. This approach has the benefit of type I (arginine to tryptophan) and Reis-Bu¨cklers’ dystrophy (ar- providing a comprehensive account of the dystrophies, which can ginine to glutamine). These studies have been confirmed and ex- be added to in the future. tended by others, and it is apparent that there are mutation hot spots at codons 124 and 555 but that mutations involving other codons are also associated with dystrophies within the same group. EPITHELIAL CORNEAL DYSTROPHIES These include mutations Arg124Leu (in GCD III, a superficial 16,17 18 Recurrent Corneal Erosion form of GCD ), Leu527Arg (in LCD IV ), and Pro501Thr 28 (LCD IIIA19). The mechanisms leading to the formation of depos- Franceschetti described recurrent corneal erosion occurring as its in these conditions is still under debate. To understand them, it a dominant trait over six generations. Onset was spontaneous or is useful to consider GCD I and LCD I as models for the kerato- was triggered by minor trauma between the ages of 4–6 years. Attacks diminished in frequency and severity by the fifth decade. epithelin dystrophies. In GCD, it has been shown that the deposits 29 contain keratoepithelin,6 together with nonkeratoepithelin pro- Other reports have also appeared. The condition is bilateral and teins, such as vimentin (an intermediate filament protein),20 cyto- characterized by microcystic epithelial lesions. Sporadic cases of keratin 18 (an epithelial protein),20 immunoglobulin (IgG) ␬ and ␭ bilateral recurrent corneal erosion are also