Linkage of a Mild Late-Onset Phenotype of Fuchs Corneal Dystrophy to a Novel Locus at 5q33.1-q35.2

Abstract

Purpose.: To identify the disease locus associated with autosomal dominant Fuchs corneal dystrophy (FCD) in a large family and to compare the progression of severity in families mapped to the FCD1 and FCD2 loci.

Methods.: Seventeen individuals in a large family were examined by slit lamp biomicroscopy. Blood samples were collected, DNA was extracted, and a genome-wide scan was performed with a microarray SNP chip. After initial generation of a genome-wide, two-point LOD score, linkage was confirmed and the critical interval was established by genotyping of short tandem repeat (STR) microsatellite markers.

Results.: A genome-wide linkage scan localized the disease interval to the long arm of chromosome 5, with a maximum two-point parametric LOD score of 3.41. Haplotype analyses refined the critical interval to 5q33.1-q35.2, spanning a 27-Mb (29-cM) region. Clinical examination of affected individuals in this family revealed an early onset of FCD at approximately age 40, after which progression of the disease was significantly attenuated compared to the FCD1- and FCD2-linked families.

Conclusions.: Late-onset FCD is linked to a novel locus on 5q33.1-q35.2 and is associated with a milder severity in age at onset and rate of progression than the FCD1 and FCD2 loci. Correlation of individual genotypes with unique rates of disease progression will provide important tools for disease management, as well as for identifying the underlying genetic lesion, offer insight into the pathomechanism of FCD.

Fuchs corneal dystrophy (FCD) is a hereditary, progressive disorder of the posterior cornea that affects up to 4% of individuals older than 40 years and represents the highest indication for corneal transplantation in the United States. ^1,2 Clinically, FCD is marked by the development of guttae, excrescences of Descemet's membrane that generally appear in the fourth or fifth decades of life and increase in number over time. ^3,4 As the disease progresses, patients present with decreased visual acuity secondary to corneal edema and endothelial cell loss, with end-stage disease evidenced by the formation of painful epithelial bullae. ⁵ This course of disease is aggravated by intraocular surgery or trauma and anterior segment inflammation, contributing to corneal decompensation and the need for corneal transplantation. ⁵  

Nearly a century after Fuchs' initial description of this phenotype, ⁶ its etiology is still not well understood. Although an early-onset FCD has been causally associated with mutations in COL8A2, ⁷ the molecular mechanisms underlying the more common late-onset form are yet to be elucidated. Previously, we reported two loci associated with late-onset FCD in large families: an FCD1 locus on chromosome 13 and a FCD2 locus on chromosome 18. ^8,9 More recently, mutations in SLC4A11, a gene associated with congenital hereditary endothelial dystrophy, ¹⁰ have been identified in a few sporadic cases of late-onset FCD. ¹¹  

Here, we report a large family with late-onset FCD. A genome-wide scan localized the disease to the long arm of chromosome 5 with suggestive LOD scores, a finding substantiated with short tandem repeat (STR) microsatellite markers. Haplotype analyses refined the critical interval to a 27 Mb (29 cM) region. Clinical examinations concluded that the severity of the disease was relatively milder, and normal progression of the disease was slower compared with the affected individuals in families linked to the FCD1 and FCD2 loci. The identification of a third FCD locus in the human genome provides another entry point for determining the genetic basis of the disorder while, for the first time, offering potential differential clinical phenotyping data that may be of value in patient management. 

Materials and Methods

Recruitment

Family members were recruited through one proband presenting to our clinic with FCD. An extended pedigree was subsequently developed through interviews, and the patients were examined in locations proximal to their area of residence. Recruitment, examination, and procedures for DNA sample collection were approved by the Institutional Review Board for Human Subjects Research at the Johns Hopkins University School of Medicine, according to the Declaration of Helsinki. Written consent was provided by each study participant, or, in the case of minors, by one of their parents. 

Determination of Phenotype and Disease Severity

Individuals underwent detailed ophthalmic examination that included slit lamp biomicroscopy. Severity was graded on a scale developed by Krachmer et al., ³ with >12 central guttae in one eye indicative of disease (grade 1), or in both eyes of individuals older than 60 years. Progression of confluence was defined by area and divided into three categories: horizontal width <2 mm (grade 2), from 2 to 5 mm (grade 3), and >5 mm (grade 4). The development of stromal and/or epithelial edema elevated this score to grade 5. Documentation included retroillumination photography performed with a digital camera (D2xs; Nikon, Tokyo, Japan) and photo slit lamp (Carl Zeiss Meditec, Inc., Dublin, CA). 

Data Analysis

All clinical severity data were entered into a spreadsheet program for statistical analyses (Excel; Microsoft, Redmond, WA). Exclusion criteria included a history of ocular inflammation, surgery, or trauma, factors that could serve as strong confounding variables. Differences in progression over time were examined by comparing clinical data with those from one family associated with the FCD1 locus and three families with the FCD2 locus. Severity was recorded according to the Krachmer scale, allowing calculation of the mean age at each level of FCD severity. We used the t-test to assess differences in severity across genotypes. 

Genotyping

A SMP microarray (ver. 5.0; Affymetrix, Santa Clara, CA) was used for genome-wide linkage analyses. Genomic DNA (500 ng) was used in a multiplexed SNP genotyping assay. Briefly, genomic DNA was digested with StyI and NspI, followed by ligation with adaptor sequences. These sequences acted as primer binding sites in PCR amplification. Amplified products were pooled and purified with magnetic beads, washed with ethanol and eluted in Tris (pH 8.0), digested into 50- to 100-bp fragments, and end-labeled by terminal deoxynucleotidyl transferase (TdT). The unincorporated labeling reagent was washed before the labeled products were loaded onto the arrays and were hybridized in an oven rotating at 60 rpm, 50°C for 20 hours. The arrays were then scanned (GeneChip Scanner 3000 7G; Affymetrix). The output signal files were checked for quality control (QC; BRLMM analysis tool 2.0 [BAT 2.0]; Affymetrix). The default threshold score was set at 0.05 and was used as a cutoff for SNPs to be excluded. Cell files were processed further to generate genotypes using a genotyping console (BAT 2.0; Affymetrix) and locally written Perl scripts. SNP genotypes were incorporated into the family pedigree files to generate PED files using Plink. ¹² Once incorporated, the same software was used to identify parentage inconsistencies. Pedigrees were analyzed assuming an autosomal dominant mode of inheritance, 0.02 disease allele frequency, and 90% penetrance by MERLIN. ¹³ SNP allele frequencies from an ethnically matched control population were obtained from HapMap (hapmap.org). 

The critical interval was delineated by genotyping STR markers. Multiplex PCR was carried out in a 25-μL mixture containing 40 ng genomic DNA, 10 μM dye-labeled primers pairs, 0.5 μL 10× PCR buffer (GeneAmp Buffer II; Affymetrix), 0.5 μL 10 mM dNTP mix, 2.5 mM MgCl₂, and 0.2 U Taq DNA polymerase (AmpliTaq Gold; Applied Biosystems, Inc. [ABI], Foster City, CA). Amplification was performed in a PCR System (GeneAmp PCR System 9700; ABI); primers and PCR conditions are available on request. PCR products were pooled and mixed with a loading cocktail containing size standards (HD-500; ABI), separated in a genetic analyzer (model 3100; ABI), and analyzed (GeneMapper software; ABI). 

Linkage Analysis

Two-point linkage analyses were performed with the FASTLINK version of MLINK from the LINKAGE Program Package (http://www.hgmp.mrc.ac.uk/ provided in the public domain by Human Genome Mapping Project Resources Centre, Cambridge, UK). Maximum LOD scores were calculated using ILINK under an autosomal dominant model with a 0.02 disease allele frequency and 90% penetrance. The marker order and distances between the markers were obtained from the Généthon database (http://www.genethon.fr/ provided in the public domain by the French Association against Myopathies, Evry, France) and the National Center for Biotechnology Information chromosome 5 sequence maps (http://www.ncbi.nlm.nih.gov/mapview/ provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD). For the initial genome scan, equal allele frequencies were assumed, whereas for fine mapping, allele frequencies were estimated from 96 unrelated and unaffected individuals. 

Results

As part of our ongoing effort to map and clone new loci for adult onset FCD, we identified a large, three-generation family, MO, through a proband examined at The Wilmer Eye Institute (Fig. 1; individual 21). On the conclusion of interviews and review of medical records, we were able to recruit and examine through slit lamp biomicroscopy a total of 17 individuals. Of these, we found 10 (eight female, two male) to fulfill the phenotypic criteria for an FCD diagnosis (12 or more central guttae in one or in both eyes of individuals above 60 years of age). The initial slit lamp examination of individual 22 detected central corneal guttae, although their number did not meet the minimum criteria for Krachmer grade 1. Hence, for the purpose of initial linkage analyses, this person was designated as unaffected. In addition, given that the father of individual 10 was found positive for FCD, we cannot rule out the possibility of her inheriting the pathogenic mutation paternally; hence, alleles of individuals 10 and 11 were not used in linkage and haplotype analyses. Last, even though the older members of family MO were certain that deceased individuals 1, 3, and 8 were positive for FCD, there were no medical records available to confirm the disease; hence, we designated the status of these individuals as unknown. 

Figure 1.

 

Pedigree of the family linked to the FCD3 locus. Filled symbols: affected; dotted symbols: trace; diagonal line through symbol: deceased. Haplotypes of 10 consecutive 5q microsatellite markers are shown, with alleles forming the disease-bearing haplotype (black).

We first asked whether this family may be linked to any of the known FCD loci. Therefore, we genotyped all available individuals with STR microsatellite markers that span the critical intervals of early- and late-onset FCD loci. Both haplotype analyses and two-point LOD scores excluded family MO from FCD1, FCD2, and the COL8A2 locus (data not shown). We therefore initiated a genome-wide scan by genotyping all appropriate family members by using an SNP microarray (5.0 platform; Affymetrix) that contains 500K SNPs spaced across the whole genome. Of the SNPs, 96% passed the initial QC and were investigated for parentage inconsistencies. A further 4560 SNPs were excluded due to non-Mendelian transmission likely attributed to bad genotyping calls, leaving us with ∼475 K SNPs. As linkage analysis (two-point or multipoint) of the entire SNP set requires exceptional computational power, we decided to simplify our dataset while maintaining uniform coverage across the genome. Hence, we selected every fifth consecutive SNP, and performed two-point linkage on ∼95 K SNPs. 

As SNPs are bi-allelic, a family consisting of 17 members would not have enough power to attain significant linkage (LOD > 3); hence, we investigated further all regions with suggestive linkage (LOD > 2). From the genome-wide scan, we observed suggestive linkage only with SNPs located on the long arm of chromosome 5. The highest LOD scores were obtained with rs13173656, rs9313417, rs2116736, rs17451810, rs4958561, and rs778816, yielding scores of 2.71, 2.68, 2.66, 2.70, 2.64, and 2.68, respectively. Reconstruction of a candidate disease haplotype localized the putative critical interval to 27 Mb on 5q. 

To determine whether this signal was truly indicative of a new FCD locus, we genotyped the region for all members of family MO with a series of 12 evenly spaced STR microsatellite markers. We observed LOD scores of 2.23, 3.35, 2.82, and 3.37 at θ = 0, with markers D5S209, D5S2093, D5S671, and D5S425, respectively (Table 1). Reconstruction of the disease haplotype was consistent with the SNP data and confirmed the linkage results further, indicating that we mapped a new locus for this disorder, FCD3. There was proximal recombination in affected individual 13 at D4S434, affected individual 14 at D5S470 and affected individual 10 at D5S397 (Fig. 1). Similarly, there was a distal recombination in affected individual 26 and unaffected individual 15 at D5S2108 and unaffected individual 20 at D5S2034 (Fig. 1). Taken together, these data place the disease locus on 5q33.1–35.2, in a 27-Mb (29-cM) region flanked by D5S470 proximally and D5S2108 distally. 

Table 1.

 

Two-Point LOD Scores for a Series of Consecutive 5q STR Markers

Marker	cM	Mb	0.00	0.01	0.05	0.10	0.30	Zmax	θmax
D5S436	147.5	145.1	−0.53	0.01	0.66	0.93	0.59	0.93	0.10
			−0.48	0.02	0.73	0.97	0.64	0.97	0.10
D5S434	150.3	147.2	−0.67	−0.19	0.51	0.77	0.45	0.77	0.10
			−0.65	−0.16	0.55	0.81	0.52	0.81	0.10
D5S470	153.2	150.2	0.81	0.95	1.12	1.10	0.50	1.15	0.07
			0.86	0.99	1.18	1.18	0.58	1.20	0.07
D5S209	156.5	152.5	2.23	2.21	2.14	2.02	0.91	2.23	0.00
			2.29	2.29	2.25	2.10	1.00	2.29	0.00
D5S2093	164.2	162.7	3.35	3.29	3.04	2.72	1.25	3.35	0.00
			3.39	3.33	3.09	2.78	1.32	3.39	0.00
D5S671	172.7	167.6	2.82	2.76	2.55	2.27	1.00	2.82	0.00
			2.88	2.82	2.60	2.32	1.05	2.88	0.00
D5S425	179.1	171.4	3.37	3.28	3.04	2.74	1.23	3.35	0.00
			3.41	3.35	3.11	2.80	1.33	3.41	0.00
D5S2108	182.4	173.5	0.55	0.79	1.10	1.15	0.56	1.14	0.10
			0.62	0.86	1.14	1.18	0.61	1.18	0.10
D5S2034	187.8	174.9	1.30	1.32	1.35	1.28	0.52	1.35	0.05
			1.33	1.35	1.37	1.30	0.59	1.37	0.05
D5S2030	190.9	177.7	−0.02	−0.01	0.01	0.02	0.02	0.02	0.40
			−0.01	−0.01	0.01	0.02	0.02	0.02	0.40

 LOD scores were calculated at different θ values for each marker with individual 22 designated as unaffected (top row) and then as affected (bottom row).

Interestingly, individual 22, scored as unaffected based on our initial inclusion and exclusion criteria, harbored the disease allele (Fig. 1), which prompted us to re-examine this person after a 2-year interval. Slit lamp examination revealed that individual 22 should be designated as clinically affected, with a Krachmer grading score of 1.5 bilaterally (Fig. 2). Because of this finding, we recalculated LOD scores both for the genome-wide set of 95 K SNPs and for the STR microsatellite markers, with individual 22 marked as affected. Not surprisingly, the two-point LOD scores improved marginally; suggestive LOD scores were only obtained with SNPs located on 5q. Similarly, LOD scores of 2.29, 3.39, 2.88, and 3.41 at θ = 0, were obtained with markers D5S209, D5S2093, D5S671, and D5S425 respectively (Table 1). 

Figure 2.

 

Photographic documentation of Fuchs corneal dystrophy associated with the FCD3 locus. Individual nonconfluent guttae are visible on slit beam and specular reflection of corneal endothelium of individuals 22 (A) and 25 (B). Of note, individual 25, a 47-year-old woman, is the daughter of individual 22, a man with the same distribution of guttae at age 73.

We were surprised at the low Krachmer score of individual 22, since he is 73 years old. Typically, older individuals of families linked to FCD1 and FCD2 loci have a comparatively severe phenotype with a Krachmer grading of 3 or higher. ¹⁴ Hence, we wondered whether there are general differences in the phenotype of the FCD3-defining family compared with the families linked to the other two FCD loci. We therefore calculated the mean age of affected individuals at each level of FCD severity and compared it with families linked to FCD1 and FCD2. The mean age of affected individuals in this family was significantly higher at Krachmer grades 1 and 2, relative to affected individuals in families linked to FCD1 and FCD2 (Fig. 3). Moreover, the average severity of affected individuals above the age of 60 was significantly less in FCD3-correlated eyes than in those associated with FCD1 (P < 0.0001) and FCD2 (P < 0.001). Among the 11 affected members of this family, two have a history of inflammatory ocular disease, in addition to FCD. Individual 21 had recurrent bilateral birdshot choroidopathy with a chronic anterior uveitis; 24 had a history of recurrent nodular scleritis and underwent argon laser trabeculoplasty for elevated intraocular pressures. As chronic anterior segment inflammation may distort the FCD phenotype, individuals 21 and 24 were not included in the severity analysis. 

Figure 3.

 

Development of disease severity over time in relation to FCD1 and FCD2. Data points represent average age per level of severity within each locus. Individuals with the disease haplotype in each family began to demonstrate early, subclinical signs of disease in the fourth decade of life, with slower progression over time evidenced in the family linked to 5q. Individuals older than 60 years in the FCD3-linked family experienced significantly reduced severity relative to those associated with the FCD1 (P < 0.0001) and FCD2 (P < 0.001) loci.

Discussion

Here, we report a new locus for late-onset FCD in a large family, with the critical interval mapped to 5q33.1–35.2. Haplotype analyses refined the interval to a 27 Mb (29 cM) region with a maximum two-point parametric LOD score of 3.41. Although the maximum LOD score of 3.41 is only slightly higher than the traditional limiting value of 3.0, it represents the maximum theoretical obtainable value for this family. Identification of the third FCD locus strongly suggests heterogeneity for late-onset FCD, contrary to early-onset that is casually associated with COL8A2. Of importance, subsequent genotyping of our unlinked FCD families capable of generating genome-wide significant linkage (n = 7) did not identify other families linked to FCD3, suggesting that FCD3, like FCD1 is not a common late-onset FCD contributor. 

Ophthalmic evaluation of the affected members of family MO suggests a mild FCD phenotype. Although advancement of disease in this family is consistent with an exponential pattern of progression, comparison with affected eyes from families linked to FCD1 and FCD2 loci reveal an initial intersection of progression curves in the fourth decade of life, which may represent a common point when subclinical signs of FCD begin to appear. Disease then progresses at unique rates based on each respective locus. At each level of severity, the mean age of individuals is increased significantly in the FCD3 family relative to FCD1 and FCD2, implying a lesser rate of disease development. Further, there is a significant difference in rates of progression between FCD1 and FCD2 families. Total comparison of these curves through all grades of severity reveals an increase that occurs on average at a younger age in members of the FCD1 family. Notably, this pattern is consistent with a previous report, in which we noted early onset in family members with disease haplotypes inherited from two different parents. ⁸ The milder phenotype noted in the present family may be a consequence of environmental and/or other genetic factors. Sequential retroillumination photography of additional families linked to FCD3 are needed to quantitatively determine the rate of progression. At the same time, it is important to identify additional FCD3-linked families to determine whether the slower progression of the phenotype is a characteristic of the locus or of the particular mutation carried by family MO. 

FCD3 maps to a gene-rich region that harbors 97 annotated genes, suggesting that traditional positional cloning approaches will be challenging. Nonetheless, the eventual identification of the genetic lesion in FCD3 will illuminate the etiopathology of FCD and potentially inform the variable expressivity and clinical progression that typifies this disorder. 

Footnotes

 Supported by National Eye Institute Grant R01EY016835 (JDG), Kwok Research Fund, and National Institute of Child Health and Development Grant R01HD04260 (NK).

Footnotes

 Disclosure: S.A . Riazuddin, None; A . O. Eghrari, None; A . Al-Saif, None; L . Davey, None; D.N . Meadows, None; N . Katsanis, None; J . D. Gottsch, None

Footnotes

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

The authors are grateful to all family members for their enthusiastic participation in this study.