The Drosophila eyes absent gene (eya) is involved in the formation of compound eyes. Flies with loss-of-function mutations of this gene develop no eyes and form the ectopic eye in the antennae and the ventral zone of the head on target expression. A highly conserved homologous gene in various invertebrates and vertebrates has been shown to function in the formation of the eye. In contrast, a human homologue, EYA1, has been identified by positional cloning as a candidate gene for branchio-oto-renal (BOR) syndrome, in which phenotypic manifestations are restricted to the areas of branchial arch, ear and kidney, with usually no anomalies in the eye. We have examined genomic DNA isolated from patients with various types of developmental eye anomaly for EYA1 mutations by the use of polymerase chain reaction–single-strand conformation polymorphism and sequencing. We identified three novel missense mutations in patients who had congenital cataracts and ocular anterior segment anomalies. One of the patients had clinical features of BOR syndrome as well. This result implies that the human EYA1 gene is also involved in eye morphogenesis, and that a wide variety of clinical manifestations may be caused by EYA1 mutations.
Received 22 September 1999; Revised and Accepted 1 December 1999.
The Drosophilaeya gene was isolated from a breakpoint region of flies with an eyeless phenotype (1). The gene has been shown to have an ability to restore the phenotype on injection of the cDNA expression construct and also to form the eye ectopically in the antennae and the ventral zone of the head on target expression (2). It has been suggested that the gene product may suppress programmed cell death in eye progenitor cells at a critical stage in eye morphogenesis (1). One of the human homologues has been isolated by positional cloning of the gene responsible for branchio-oto-renal (BOR) syndrome (3), and subsequent studies have revealed that the homologues designated as Eya1, Eya2 and Eya3 form a novel gene family in humans and mice (4,5). The original report demonstrated by in situ hybridization that murine Eya1 was expressed in the otic, olfactory and renal primordia, but not in the branchial arch and eye (3). However, a subsequent report showed that the EYA family members were also expressed in the developing eye: EYA1 was expressed in the ocular anterior segment, lens and optic nerve sheath, and EYA2 in the posterior segment (5). Thus, we speculated that the EYA family might be a potential candidate for eye anomalies; however, no anomalies in the eye have been described in patients with BOR syndrome. In this study, we have detected three novel missense mutations of the EYA1 gene in patients with developmental ocular anomalies. The mutations provide an interesting genotype–phenotype correlation.
Clinical phenotypes of patients carrying the mutation
We have screened for EYA1 mutations in genomic DNA isolated from patients with various types of eye anomaly (a total of 317 samples as described in Materials and Methods), and detected three mutations in the following patients.
Patient 1 is a 4-year-old girl who was introduced to our hospital with visual impairment. Ocular examinations revealed central corneal opacity, adhesion to the iris (Peters’ anomaly) and slight cataracts in both eyes, whereas the fundus was normal (Fig. 1a). Her visual acuity was 0.08 bilaterally (the score is a fraction of the visual angle with which two points can be recognized separately; visual acuity is usually measured using a chart with the Landolt rings and a normal score ranges from 1.0 to 2.0). Her mother, aged 32, had nuclear-type congenital cataracts with 0.3 visual acuity. The patient and her mother were otherwise normal in appearance, intelligence and karyotype (46,XX). After finding the mutation, we asked our colleagues in the Departments of Otolaryngology and Pediatrics (National Children’s Hospital, Tokyo, Japan) to conduct careful examinations of the patient, but no clinical findings suggesting BOR syndrome were obtained except for a slight elevation of the auditory brain stem response (ABR) threshold in hearing.
Patient 2 is a 3-year-old boy who presented at our hospital with iris anomaly. Eye examinations revealed bilateral persistence of the pupillary membrane, but a normal lens and fundus (Fig. 1b). His visual acuity was 0.5 in the right eye and 0.8 in the left eye with correction. He had no other systemic abnormalities, and was found to be normal in growth and intelligence for his age, even after careful re-examinations following the finding of the mutation as described for patient 1.
Patient 3 is an 8-year-old boy who had first presented at our hospital with nystagmus and systemic edema at 20 days of age. Examinations revealed bilateral nuclear-type congenital cataracts with the normal fundus, and multicystic dysplasia in his right kidney, which did not function and caused hypocalcaemia (Fig. 1c and d). The cataracts were operated on at 1 month of age and the right kidney was removed at 2 months. He was later diagnosed as having conductive deafness with the malleus anomaly. He also has cervical fistula that occluded spontaneously. These clinical findings, besides the cataracts, coincide with typical features of BOR syndrome. He now has 0.2 visual acuity with esotropia, and slight mental retardation.
Mutations of the EYA1 gene
Patient 1 and her mother had an A→G nucleotide substitution at position 1688 of a cDNA form of the EYA1 gene (GenBank accession no. Y10260, in exon 15), which is expected to result in R514G (Fig. 2a). Patient 2 had a G→A substitution at position 1136 (exon 10), which resulted in E330K (Fig. 2b). The mutation was not detected in his parents, who are apparently normal, thus indicating that it was sporadic. Patient 3, but not his normal parents, had a G→A substitution at position 1325 (exon 12), which results in G393S (Fig. 2c).
All the mutations detected above occurred on one of the alleles (heterozygous) and were not detected in unaffected members of the immediate family nor in >100 normal individuals. The relationship of biological paternity and maternity in the above pedigrees was confirmed with multiple microsatellite markers.
BOR syndrome (OMIM 113650) is an autosomal dominant disorder with incomplete penetrance and variable expressivity. Anomalies are usually detected in the latero-cervical fistulas, in the outer, middle and inner ear, and in the kidney. However, BOR patients show different combinations of these symptoms with varying degrees of severity even within the same family. Branchio-otic (BO) syndrome (OMIM 602588), which had been considered to be a different medical entity because of the absence of renal anomaly, has been demonstrated to be allelic with the finding of mutations of the EYA1 gene (6). Most mutations identified to date in the BOR and BO syndromes are situated in the C-terminal region (271 amino acid residues encoded by the last eight exons, the eyaHR region), and cause translational termination by nonsense and frameshift mutations as well as by splice errors (3,4,6,7). These mutations occur on one of the alleles, thus haploinsufficiency of the gene product has been suggested to cause the BOR phenotype. A recent study on heterozygous Eya1 mice generated with knockout technology supports this (8). In contrast, only two missense mutations, S454P and L472R, have been identified (3,4).
There has been no description of unusual eyes for BOR patients, although anomalies of the anterior segments of the eye are easily recognized. Anomalies of the lacrimal ducts are sometimes associated with BOR syndrome (9), but the lacrimal ducts differ from the structure of the eye in terms of the morphogenic pathway (10). We have detected three novel missense mutations in patients with anomalies in the anterior segments of the eye. The mutations detected occurred in the eyaHR region, which is relatively conserved among EYA family members of different species (3–5,11,12). Notably, the glutamic acid residue at 330 and the glycine residue at 393 are conserved among all the EYA family members identified to date, whereas the arginine residue at 514 is conserved in EYA1 and EYA2 of both humans and mice, and lysine occurs in EYA3 (1,3–5). These amino acid residues may play an important role in its reaction or maintenance of the protein structure, but molecular studies in experimental systems are required to reveal the function. Currently, there are no experimental systems with which to investigate its function, although the EYA1 product is suggested to be involved in apoptosis.
Patient 3 had cataracts in addition to renal and otic anomalies, thus he is an atypical BOR syndrome patient. In contrast, the affected areas are almost limited to the eye in the other two patients, which indicates that some forms of the EYA1 mutations seem to affect only the morphogenesis of the eye and not the branchial arch or ear. It is interesting that the three novel mutations detected in this study are missense mutations and do not result in premature termination of translation. Therefore, the mutated protein with an amino acid substitution may have a dominant effect especially in the process of eye morphogenesis. Peters’ anomaly derives from insufficient separation of the lens vesicle from the surface ectoderm, and persistent pupillary membrane represents an incomplete involution of anterior tunica vasculosa lentis (13). Thus, these diseases may be accounted for by suppression of programmed cell death in progenitor cells of the anterior segment of the eye at an early stage, and probably also in lens fiber cells in cataracts.
Two of the detected mutations do not exist in the biological parents. This high incidence of sporadic mutation may be accounted for by our sample collection. Our pooled specimens are derived from an affected individual and are not intended to cover large affected families. Although we hear of the family history from the patient in clinical practice, we only examine other family members directly when a mutation is detected. The other mutation is associated with affected individuals in two generations, but the evidence is not so strong that the mutation causes the disease. However, as EYA family members of other species are clearly involved in eye morphogenesis, our study recaptures the attention paid to ocular anomalies caused by the EYA1 gene, which was once thought not to relate to the eye of humans. Extensive surveys for mutations, especially in large affected families, and functional analyses of the gene product would validate the significance of the mutation we detected.
MATERIALS AND METHODS
This study was conducted in accordance with the World Medical Association Declaration of Helsinki. Our use of human subjects was conducted under the program approved by the National Children’s Hospital Experimental Review Board, and deemed exempt from human subject regulations. We collected DNA samples from individuals with various types of congenital eye anomaly. The samples included 53 patients with ocular anterior segment anomalies including Peter’s anomaly, 43 patients with aniridia, 71 patients with congenital cataracts, 22 patients with isolated foveal hypoplasia and 128 patients with optic nerve anomalies, and were used in analyses for PAX6 mutations (14–18). DNA samples of the family members were also collected when a mutation was detected. All the subjects in our study are apparently Japanese. After obtaining informed consent, blood samples were collected from peripheral veins into lithium heparin tubes. Genomic DNA was prepared from isolated leukocytes using a standard phenol–chloroform procedure. The DNA samples used for normal controls have been described previously (19,20).
Polymerase chain reaction–single-strand conformation polymorphism (PCR–SSCP) assay and sequencing
PCR primers used for amplification of exons 1–16 of the EYA1 gene were synthesized using a DNA/RNA synthesizer (model 392; Applied Biosystems, Foster City, CA) based on previous reports (3,4). PCR was performed with 100 ng of genomic DNA and rTaq DNA polymerase (Takara, Shiga, Japan) in PCR buffer containing 1.5 mM of MgCl2 and [α-32P]dATP for 30 cycles of 94°C for 1 min, 55°C for 1 min and 72°C for 3 min. The annealing temperature was adjusted to 45°C for exon 10. Products were denatured at 94°C for 5 min and loaded onto 5% non-denatured polyacrylamide gels. The following four running buffers were used: 0.5× TBE with or without 10% glycerol and 1.0× TBE with or without 10% glycerol. The running condition was either 4°C or room temperature. Thus, we analysed the products in eight different conditions for SSCP. After running for 12–18 h at 200–300 V, gels were dried and exposed to an X-ray film for 6–24 h. PCR products showing an aberrant mobility compared with those of healthy donors were sequenced after subcloning on pUC18 using a DNA sequencing kit (Amersham, Cleveland, OH) and an automatic DNA sequencer (A373; Applied Biosystems). Sequence variations were confirmed in at least six independent colonies, and also with direct sequencing of the amplified products.
This study was supported in part by grants for Eye and Ear Science Research, for Pediatric Research and for Human Genome Research from the Ministry of Health and Welfare, Japan, and a grant of Organized Research Combination System from the Science and Technology Agency, Japan.
To whom correspondence should be addressed. Tel: +81 3 3414 8121 ext. 2197; Fax: +81 3 3419 0381; Email: firstname.lastname@example.org