Four types of autosomal dominant corneal dystrophy, including

Transcription

1 Severe Corneal Dystrophy Phenotype Caused by Homozygous R24H Keratoepithelin Mutations Masaki Okada, Shuji Yamamoto, Yoshitsugu Inoue, l Hitoshi Watanabe, x Naoyuki Maeda, l Yoshikazu Shimomura, l Yasuo Ishii, 2 and Yasuo Tano A PURPOSE. TO determine the mutational status of the fiig-h3 gene in five patients from four Japanese families affected with an unusual, severe form of corneal dystrophy. In these five cases, the corneas were remarkable for confluent round opacities in the superficial stromal layer. The fiig-h3 gene coding for keratoepithelin was recently identified as the gene responsible for 5q-linked autosomal dominant corneal dystrophies. METHODS. Genomic DNA was isolated from leukocytes of five patients with the severe form of corneal dystrophy. To screen for point mutations, exons of the (5ig-h3 gene were amplified by polymerase chain reaction and were analyzed with the single-strand conformational polymorphism technique. Subsequently, the mutations were identified by a direct sequencing method and restriction enzyme digestion analysis. RESULTS. All five patients with the severe form of corneal dystrophy had homozygous R24H keratoepithelin mutations. Histopathologic examinations of the corneas obtained from two patients with the severe form showed granular, rod-shaped deposits. CONCLUSIONS. The severe phenotype was a pathologic variant of granular corneal dystrophy (GCD). All five patients had homozygous R24H keratoepithelin mutations. The R24H keratoepithelin mutation is the same mutation recently reported to be responsible for Avellino corneal dystrophy. The homozygous R24H keratoepithelin mutations are the cause of the severe variant of GCD characterized by juvenile-onset and confluent superficial opacity. {Invest Ophthalmol Vis Sci. 998;39: ) From the ' Department of Ophthalmology, Osaka University Medical School, and the department of Ophthalmology, Shinkawabashi General Hospital, Japan. Supported in part by an unrestricted grant for Research on Disability Health and Welfare from the Ministry of Health and Welfare, Tokyo,Japan. Submitted for publication December 2, 997; revised March 6, 998; accepted May 22, 998. Proprietary interest category: N. Reprint requests: Shuji Yamamoto, Department of Ophthalmology, Osaka University Medical School (E7), 2-2 Yamadaoka, Suita, Osaka , Japan. Four types of autosomal dominant corneal dystrophy, including granular corneal dystrophy (GCD), lattice corneal dystrophy type I (LCDI), Avellino corneal dystrophy (ACD),'~? and Reis-Bucklers' corneal dystrophy, have been genetically mapped to the same region of 5q3 4 Recently, Munier et al. 5 constructed a yeast artificial chromosome contiguous map covering the region where these four autosomal dominant corneal dystrophies were mapped and isolated the {tig-b3 gene. The \Sig-h3 gene, inducible by transforming growth factor-/3, was first cloned by Skonier et al. 6 Munier et al. considered this gene to be a good candidate gene for corneal dystrophy because it also maps to 5q3 and is expressed almost exclusively in the corneal epithelium. They therefore initiated a systematic search for point mutations in the (3ig-h3 gene in people with these four types of corneal dystrophy. Four types of heterozygous mutations were detected in two arginine codons in the gene, codon 555 and codon 24. R555Wwas found in a family with GCD, R555Q in a family with Reis-Biicklers' corneal dystrophy, R24C in two families with LCDI, and R24H in two families with ACD. 5 Keratoepithelin is a synonym for the fiig-h3 gene. We studied five patients who were from four Japanese families and who were affected with an unusual, severe form of corneal dystrophy. Three of these five people were offspring of consanguineous marriages. Clinical examinations of other family members showed that several members were affected with a mild form of corneal dystrophy. The mild form seen in these four families had clinical features of ACD. Moreover, the mild form had an autosomal dominant mode of inheritance. The intrafamilial phenotypic variability observed in these four pedigrees prompted us to analyze the associated genomic fiig-h3 status for possible genotypephenotype correlation. METHODS Patients Table shows the clinical data of the five patients who were from four Japanese families and were affected with an unusual severe form of corneal dystrophy (patient numbers: K-3, K-5, S-l, O-l, and M-l). They were aged 30 to 47 years; age at onset of symptoms ranged from 4 to 6 years. Patients K-3 and K-5 were siblings. The pedigrees of these four families are shown in Figure. Consanguineous marriages were found in two pedigrees (K and S). Four patients, except Patient O-l, were offspring of parents both of whom had a mild form of corneal dystrophy. Dominant inheritance of the mild phenotype was documented in all families, except in family S in which transmission of the mild phenotype by the two affected parents was not observed (Fig. ). Thesefivepatients had undergone one to five lamellar keratoplasties each, because of early and progressive recurrence of the disease. The corneas of these five patients were remarkable for confluent round white opacities in the superficial stromal layer (Fig. 2A). Clinical examinations of family members with the mild phenotype showed fewer round opacities and several star-shaped, spicular snowflakelike opac- Investigative Ophthalmology & Visual Science, September 998, Vol. 39, No. 0 Copyright Association for Research in Vision and Ophthalmology 947

2 948 Reports IOVS, September 998, Vol. 39, No. 0 TABLE. Clinical Characteristics of Patients Patients Sex Age Age at Onset CM VA (OD) VA (OS) ST (OD) ST (OS) K-3 K-5 S-l O-l M- F F F M M Yes Yes Yes No No 20/ /00 20/00 20/40 20/40 20/200 20/50 20/200 20/00 20/ CM, consanguineous marriage; VA, best corrected visual acuity; ST, number of surgical treatments. Family K II III IV 4T Family S K-7 K-8 Family O II D s-i.. o-i Family M /^ /^ II III M- FIGURE. Pedigrees of the five patients with a severe form of corneal dystrophy. *A patient with a severe form. Solid symbols indicate family members with a mild form of corneal dystrophy. Open symbols indicate unaffected family members. Double lines indicate consanguineous marriages.

3 IOVS, September 998, Vol. 39, No. 0 Reports 949 FIGURE 2. Clinical photographs and microscopic photographs. (A) The right eye of the patient with a severe form of corneal dystrophy (patient M-l) had confluent discrete round opacities. (B) The right eye of the patient with a mild form (father of M-l) had fewer round opacities and several star-shaped spicular opacities. (C) Electron microscopy of the cornea obtained from the patient with a severe form (patient K-5) showed that rod-shaped dense materials (R) had accumulated. The keratocyte (K) was degenerated (asterisks). Glycogen granules were seen in the keratocyte (arrow). (C, insets) Deposits stained with Luxol fast blue (top left) and toiuidine blue (bottom right) had accumulated in the superficial stromal layer (arrows). Ep, epithelium; St, stroma. (continued on next page) ities, which are consistent with clinical features of ACD (Fig. 2B). 3 The prominent lattice lesions seen in lattice corneal dystrophy type I were not detected by slit lamp examination in any patients with the severe or mild phenotype. Histologic Examinations Using light and electron microscopy, we examined the corneal buttons obtained after grafting procedures on two patients affected with the severe form of corneal dystrophy (K-3 at age

4 IOVS, September 998, Vol. 39, No. 0 FIGURE 2. (Continued) 28 and K-5 at age 30). The comeal buttons were bisected immediately after grafting. Half of each button was fixed in 4% formaldehyde, embedded in paraffin, and sectioned for histologic studies. The sections were stained with hematoxylin and eosin, toluidine blue, Luxol fast blue, Masson trichrome, and Congo red. The other half of each button was fixed in 4% glutaraldehyde, postfixed in % osmium tetroxide, embedded in epoxy resin, and sectioned for transmission electron microscopy. Molecular Analysis Molecular analysis was performed on all five patients affected with the severe form of corneal dystrophy. Five other family members affected with the mild form of corneal dystrophy (K-l, K-2, K-4, K-6, and K-7), and one unaffected family member (K-8) were also analyzed. We collected 20 ml venous blood from each participant and extracted genomic DNA. The singlestrand conforniational polymorphism technique was used to screen for point mutations. Genomic DNA was amplified by polymerase chain reaction with 500 nm each forward and reverse primer in an amplification mixture containing 0 mm Tris-HCI (ph 8.3), 50 mm KC,.5 mm MgCl 2, 0.2 mm deoxyribonucleoside triphosphate, and 0.5 U Taq polymerase (Ampli- Taq Gold, Perkin-Elmer, Branchville, NJ). The primer pairs, which were based on published sequences, 5 amplified each exon of the fiig-b3 gene, including flanking introns. Amplified DNA was mixed with an equal volume of formamide dye. Denatured samples were subjected to electrophoresis on 7% polyacrylamide gel containing 0% glycerol for 2 to 20 hours at 4 W at room temperature. After electrophoresis, the gels were stained with a nucleic acid gel stain solution (SYBR green; Molecular Probes, Eugene, OR) and analyzed using a laser scanning image analyzer CFluor Imager 595; Molecular Dynamics, Sunnyvale, CA). Exons that displayed abnormal bands were sequenced using a direct sequencing method. Amplified DNA was purified with a polynierase chain reaction purification kit (Qiagen, Hilden, Germany) and sequenced with an automatic fluorescent DNA sequencer (ABI Prism 377; Applied Biosystems, Foster City, CA) and a dye terminator cycle sequencing kit (Perkin-Elmer). Restriction enzyme digestion analysis was performed to screen for the mutations detected by single-strand conforniational polymorphism and direct sequencing. Primer pairs, sense 5'-CCCCAGAGGCCATCCCTCCT-3' and antisense 5'- TCAGGCCTCAGCTTCACCG-3', were used to amplify exon 4 from the exon-intron boundary to position 439- Subsequently, the polymerase chain reaction product was digested with a mutation-specific restriction enzyme, Dralll (BioLabs, Beverly, MA). Electrophoresis was performed on a 4% polyacrylamide gel, which was stained with the same nucleic acid gel stain solution (SYBR green; Molecular Probes) solution

5 IOVS, September 998, Vol. 39, No. 0 Reports 95 A: normal A C G G A C C G C A C G G A G B: K- C: K-3 FIGURE 3- Direct sequencing of exon 4 of the fiig-h3 gene. (A) Sequence of a normal person showed CGC nucleotides in codon 24 that code for an arginine. (B) Sequence of patient K- (mild case) showed double peaks of G (black) and A (green) at the second nucleotide position in codon 24, by which the second nucleotide could not be distinguished (N, not distinguished). This G to A transition results in a substitution from arginine to histidine in codon 24 (Argl24His). (C) Sequence of patient K-3 (severe case) showed CAC nucleotides in codon 24. He had the homozygous Argl24His mutations. mentioned earlier. DralU recognizes the sequence, 5'-CAC- NNN/GTG-3'. This study was conducted according to the tenets of the Declaration of Helsinki and was approved by our Institutional Human Experimentation Committee. Informed consent was obtained from all participants. RESULTS Light microscopy of the corneas obtained from the two patients (K-3 and K-5) affected with the severe form of corneal dystrophy showed deposits in the superficial stromal layer that stained with Luxol fast blue (Fig. 2C-inset, top left) and Masson trichrome (data not shown). The deep stromal layers were normal. None of the stromal layers was stained with Congo red (data not shown). Electron microscopy showed an accumulation of rod-shaped, dense materials (Fig. 2C). These pathologic findings agreed with the characteristic deposits seen in GCD, ACD, and Reis-Biicklers' corneal dystrophy. The single-strand conformational polymorphism analysis of the severely affected patient (K-3) showed an abnormal

6 952 Reports IOVS, September 998, Vol. 39, No. 0 M N K K2 K3 K4 K5 K6 K7 K8 S O M 200bp 00bp 46 wild type Arg24His FIGURE 4. Restriction enzyme digestion analysis specific to the Argl24His keratoepithelin mutation. Upper bands (40 bp) represent the wild type. Lower digested bands (approximately 20 bp) represent the Argl24His mutation, Unaffected members (N, K-8) did not have the mutation. Family members with a mild form (K-l, K-2, K-4, K-6, and K-7) had the mutation heterozygously. All five patients with a severe form (K-3, K-5, S-l, 0-, and M-l) had the mutation homozygously. M, 00-bp ladder marker; N, normal person. band in exon 4 of the fiig-h3 gene. No abnormal band was detected in other exons (data not shown). Direct sequencing of exon 4 of the $ig~h3 gene showed that patient K-3 had homozygous R24H mutations (from CGC to CAC in codon 24). Her father (K-l) had a heterozygous R24H mutation (Fig. 3). Restriction enzyme digestion analysis was performed to screen for the RI24H mutation. All five severely affected patients had the homozygous R24H mutations. The other family members affected with the mild phenotype had the heterozygous R24H mutation, and the unaffected members (a normal person and K-8) were carriers of the wild-type sequence (Fig. 4). DISCUSSION We report five patients affected with an unusual severe form of corneal dystrophy from four Japanese families. The clinical manifestations in these five patients were confluent round opacities in the superficial stromal layer. The onset was at an early age, and the recurrence after surgical treatments was early and progressive. Pathologic findings in the corneas obtained from two patients displayed granular deposits, which were the main deposits seen in GCD and ACD.' In these four pedigrees, a mild phenotype of corneal dystrophy had an autosomal dominant mode of inheritance. Unfortunately, we could not examine their corneas histopathologically, because most of the family members with the mild phenotype were subclinical. Therefore, the pathologic diagnosis of the mild cases remained unclear. The severe phenotype of our cases is reminiscent of the superficial juvenile granular dystrophy previously reported by several investigators. 7 " 9 Superficial juvenile granular dystrophy is an unusual variant of GCD. All reported patients with this variant were of Asian origin. In addition, the clinical manifestations and pathologic findings of this variant are similar to those in our severe cases, except that Owens et al. 9 have reported one distinctive case associated with trace amounts of amyloid deposition. Our molecular analysis showed that all five patients with the severe phenotype had homozygous R24H keratoepithelin mutations. Moreover, other family members affected with the mild phenotype had a heterozygous R24H mutation. The R24H keratoepithelin mutation has recently been reported to be responsible for ACD. 5 The severe phenotype is caused by the genotype of the homozygous R24H keratoepithelin mutations responsible for ACD. Avellino corneal dystrophy is a variant of GCD associated with amyloid deposition. " 3 We surmise that the case reported by Owens et al. y probably had a homozygous genotype for ACD, because amyloid deposits were certainly present. However, no amyloid deposition was detected pathologically in the corneas derived from our two homozygous patients. It is possible that minute amyloid deposits were present in other areas of the corneal specimen or that amyloid deposits were masked by the hyaline deposits. We have recently found another severe variant of GCD with the homozygous R555W keratoepithelin mutations responsible for typical GCD. 0 Its phenotype of continuous placoid opacification is distinct from the phenotype of confluent discrete opacities caused by the homozygous R24H mutations. It is impossible for us to study the genotype of the cases previously reported by Rodrigues et al. 7 and Sajjadi and Javadi. 8 However, the corneal dystrophy reported in these Asian cases was clinically similar to that in our five severe cases: There was no amyloid deposition, as was true in our two homozygous cases. From these observations, we can draw two conclusions. First, in our five Japanese cases, the severe variant of GCD characterized by juvenile onset and confluent discrete opacities are caused by the genotype of homozygous R24H keratoepithelin mutations responsible for ACD. A similar phenotype of superficial juvenile granular dystrophy previously reported by several investigators is probably caused by the same genotype. Second, the mechanism that rendered the corneas of our homozygous patients devoid of amyloid deposits is unknown. The amyloid component appears later in life in ACD. 2 ' 3 Typical amyloid lesions in ACD are seen after age 40 and in the deep stromal layer. However, our homozygous patients are relatively young, and corneal opacities in homozygous cases are mostly subepithelial. We suggest that the variable expressivity of amyloid deposition between heterozygous

7 IOVS, September 998, Vol. 39, No. 0 Reports 953 and homozygous ACD patients is caused by the difference in age and accumulated layers of deposition. Previously, corneal dystrophies have been classified by clinical and histopathologic findings. Therefore, differential diagnosis of corneal dystrophy is complicated. Our results suggest that in the future, corneal dystrophies should be classified according to their genotype. References. Folberg R, Alfonso E, Croxatto JO, et al. Clinically atypical granular corneal dystrophy with pathologic features of lattice-like amyloid deposits: a study of three families. Ophthalmology. 988;95: Holland EJ, Daya SM, Stone EM, et al. Avellino corneal dystrophy: clinical manifestations and natural history. Ophthalmology. 992; 99: Rosenwasser GOD, Sucheski BM, Rosa N, et al. Phenotypic variation in combined granular-lattice (Avellino) corneal dystrophy. Arch Ophthalmol. 993; : Stone EM, Mathers WD, Rosenwasser GOD, et al. Three autosomal dominant corneal dystrophies map to chromosome 5q. Nat Genet. 994;6: Munier FL, Korvatska E, Djemai A, et al. keratoepithelin mutations in four 5q3-linked corneal dystrophies. Nat Genet. 997;5: Skonier J, Neubauer M, Madisen L, et al. cdna cloning and sequence analysis of beta-ig-h.3, a novel gene induced in a human adenocarcinoma cell line after treatment with transforming growth factor-beta. DNA Cell Biol. 992;ll: Rodrigues MM, Gaster RN, Pratt MV. Unusual superficial confluent form of granular corneal dystrophy. Ophthalmology. 983;90: Sajjadi SH, Javadi MA. Superficial juvenile granular dystrophy. Ophthalmology. 992;99: Owens SL, Sugar J, Edward DP. Superficial granular corneal dystrophy with amyloid deposits. Arch Ophthalmol. 992;! 0: Okada M, Yamamoto S, Watanabe H, et al. Granular corneal dystrophy with homozygous mutations in the keratoepithelin gene. Am] Ophthalmol. In press. Cytokine Expression during Orthotopic Corneal Allograft Rejection in Mice Yoichiro Sano, Hideya Osawa, Chie Sotozono, and Shigeru Kinoshita PURPOSE. The acquisition of cell-mediated immunity against donor antigens has been shown to be associated with rejection of orthotopic corneal allografts, but the mechanisms that cause corneal allograft destruction in grafted tissue remain obscure. To determine which T-cell subsets infiltrate graft tissue and cause graft rejection, cytokine expression was examined in corneal tissue after orthotopic corneal allograft. METHODS. BALB/C mice received orthotopic corneal allografts from either syngeneic BALB/c or allogeneic C57BL/6 donors. At or 4 weeks after grafting, the mice were euthanatized, and their corneas were removed. Corneal tissue was frozen, homogenized, and placed in phosphate-buffered saline (PBS). Each sample consisted of five corneas in 500 ml PBS. After centrifugation, the supernatant was collected, and the concentration of the following cytokines was measured by enzyme-linked immunosorbent assay: interleukin (LL)-la, IL-2, IL-4, IL-0, interferon 0FN)-y, and tumor necrosis factor (TNF)-a. From the Department of Ophthalmology, Kyoto Prefectural University of Medicine, Kyoto, Japan. Supported in part by Research Grant from the Japanese Ministry of Education, Culture and Science and a research fund from Kyoto Foundation for the Promotion of Medical Science. Submitted for publication November 20, 997; revised March 30, 998; accepted April 5, 998. Proprietary interest category: N. Reprint requests: Yoichiro Sano, Department of Ophthalmology, Kyoto Prefectural University of Medicine, Kajii-cho, Kawaramachi-Hirokoji, Kamigyo-ku, Kyoto City, Japan 602. RESULTS. Significantly increased amounts of proinflammatory cytokines (LL-la and TNF-a) were detected in supematants from all grafted corneas (both syngeneic and allogeneic) at week after grafting. At 4 weeks after grafting, supematants from normal corneas, corneas with syngeneic grafts, and corneas with accepted corneal allografts contained undetectable amounts of IL-2 and LFN-y, whereas supematants from corneas with rejected corneal allografts contained significant amounts of IL-2 and IFN-y. There were no significant differences in the amounts of IL-4 or IL-0 among all samples. Histologic examination confirmed the expression of IL-2 and IFN-y in rejected corneal allografts. CONCLUSIONS. Because IL-2 and IFN-y are secreted primarily by T-helper type (Th ) cells, whereas IL-4 and IL-0 are secreted by T-helper type 2 (Th 2) cells, these results indicate that Th -type cytokines, rather than Th 2-type cytokines, contribute to the rejection of orthotopic corneal allografts in graft tissue. (Invest Ophthalmol Vis Set. 998;39: ) tudy of the immune response in orthotopic corneal trans- has revealed unique features associated with the Splantation anterior segment of the eye, compared with other types of organ transplantation. Unlike other types of organ transplants, corneal grafts that display major histocompatibility complex class I or class II alloantigens only are less likely to be rejected, whereas grafts that display only minor transplantation antigens are rejected much more frequently.' >2 It has been shown also that the acceptance of orthotopic murine corneal allografts correlates positively with the development of donor-specific anterior chamber-associated immune deviation. 3 Moreover, recent studies have reported that donor-specific delayed hypersensitivity responses detected after orthotopic corneal allografts are directed at donor-minor alloantigens, but not at major histocompatibility complex alloantigens. 4 Using experimental animals, Callanan et al 5 performed a study to characterize the mechanisms of orthotopic corneal allograft rejection and reported a correlation between rejection and cell-mediated immunity in the recipients. Although this study revealed the acquisition of systemic cell-mediated immu-