Vision 2012; 18:1449-1456
Received 18 October 2011 | Accepted 30 May 2012 | Published 2 June 2012
1Jules Gonin Eye Hospital, Lausanne, University of Lausanne, Lausanne, Switzerland; 2Department of Radiology, CHUV, Lausanne, University of Lausanne, Lausanne, Switzerland; 3Genetics Unit, Department of Pediatrics, University of Alexandria, Alexandria, Egypt; 4Department of Ophthalmology, University of Alexandria, Alexandria, Egypt; 5IRO - Institute for Research in Ophthalmology, Sion, Switzerland; 6EPFL - Ecole polytechnique fédérale de Lausanne, Switzerland Faculty of Life Sciences, Lausanne, Switzerland
Correspondence to: Hana Abouzeid, Jules Gonin Eye Hospital, 15 Avenue de France CH-1004 Lausanne, Switzerland; Phone: +4121626 8592; FAX: +4121626 8888; email: email@example.com
Purpose: To report the clinical and genetic study of two families of Egyptian origin with clinical anophthalmia. To further determine the role of the retina and anterior neural fold homeobox gene (RAX) in anophthalmia and associated cerebral malformations.
Methods: Three patients with clinical anophthalmia and first-degree relatives from two consanguineous families of Egyptian origin underwent full ophthalmologic, general and neurologic examination, and blood tests. Cerebral magnetic resonance imaging (MRI) was performed in the index cases of both families. Genomic DNA was prepared from venous leukocytes, and direct sequencing of all the exons and intron-exon junctions of RAX was performed after PCR amplification.
Results: Clinical bilateral anophthalmia was observed in all three patients. General and neurologic examinations were normal; obesity and delay in psychomotor development were observed in the isolated case. Orbital MRI showed a hypoplastic orbit with present but rudimentary extraocular muscles and normal lacrimal glands. Cerebral MRI showed agenesis of the optic nerves, optic tracts, and optic chiasma. In the index case of family A, the absence of the frontal and sphenoidal sinuses was also noted. In the index case of family B, only the sphenoidal sinus was absent, and there was significant cortical atrophy. The three patients carried a novel homozygous c.543+3A>G mutation (IVS2+3A>G) in RAX. Parents were healthy heterozygous carriers. No mutations were detected in orthodenticle homeobox 2 (OTX2), ventral anterior homeobox 1 (VAX1), or sex determining region Y-box 2 (SOX2).
Conclusions: This is the first report of a homozygous splicing RAX mutation associated with autosomal recessive bilateral anophthalmia. To our knowledge, only two isolated cases of anophthalmia, three null and one missense case affecting nuclear localization or the DNA-binding homeodomain, have been found to be caused by compound heterozygote RAX mutations. A novel missense RAX mutation was identified in three patients with bilateral anophthalmia and a distinct systemic and neurologic phenotype. The mutation potentially affects splicing of the last exon and is thought to result in a protein that has an aberrant homeodomain and no paired-tail domain. Functional consequences of this change still need to be characterized.
Anophthalmia (absence of the eye) is rare and the most severe ocular dysgenesis. Often discussed together with microphthalmia (small eye), anophthalmia and microphthalmia have a cumulative approximate frequency of one to two in 10,000 births [1-3]. Despite this low rate, the severity of the disease and the role it can play in understanding normal eye development justify thorough study of human anophthalmia.
Several anophthalmia- or microphthalmia-causing genes have been identified to date and include, among others, paired box 6 (PAX6), orthodenticle homeobox 2 (OTX2), sex determining region Y-box 2 (SOX2), visual system homeobox 2 (VSX2), ventral anterior homeobox 1 (VAX1), and retina and anterior neural fold homeobox (RAX) [4-6]. SOX2 is to date the gene most frequently involved in anophthalmia and accounts for up to 10% of cases [5,7].
RAX is a homeobox gene that plays a major role in human and vertebrate eye development. In mice, Rax is involved in optic vesicle formation, and loss of function is responsible for anophthalmia and leads to brain malformation . Expressed early in retinal development, RAX is thought to determine the fate and proliferation of retinal cells . Few reports of RAX mutations in humans have been published, and these include anophthalmia, microphthalmia, and eye coloboma [10,11].
This study was approved by the Ethics Committee of the Faculty of Medicine of the University of Alexandria, Egypt, and was conducted in accordance with the tenets of the Declaration of Helsinki. Written informed consent was obtained from each participant or parent. Three patients with anophthalmia belonging to two consanguineous families and their first-degree relatives were included in this study (Figure 1: family A, patients IV:1 and IV:5; family B, the affected daughter). Parents of the affected patients were first-degree cousins in both families. In family A (Figure 1), two affected children died in infancy because of severe dehydration caused by diarrhea (IV:3, IV:4). The two families were from the District of Alexandria in northwestern Egypt. All available subjects underwent full ophthalmic, general, and neurologic examinations at the Departments of Ophthalmology, Pediatrics, and Neurology of the University of Alexandria, Egypt, respectively. Special attention was paid to assessing the presence of associated anomalies such as mental retardation. A cerebral magnetic resonance imaging (MRI) was performed on both index patients (patient IV:5 of family A and the affected daughter of family B; Figure 1). Magnetic resonance imaging (MRI) were reviewed by two different neuroradiologists. MRI acquisition techniques included conventional T1- and T2-weighted multisection images (5 mm slice) on a 1.5-Tesla MR imaging system (Siemens, Helsinki, Finland).
Microsatellite markers flanking all known genes associated with micro/anophthalmia were evaluated for homozygosity in families A and B. DNA from patients was extracted from peripheral leucocytes as previously described . Direct bidirectional sequencing of all PCR-amplified exons including the 3′- and 5′ untranslated regions as well as the adjacent intron junctions was performed with the ABI Dye Terminator (Applied Biosystems, Zug, Switzerland), version 1, and electrophoresed on a 3130XL ABI Genetic Analyzer (Applied Biosystems). Primers were obtained with the help of the Primer3 program and are listed in Table 1. Sequences were compared to the reference sequence NM_000280.3 using Chromas version 2.23 (Technelysium, Tewantin, Australia). The adenine of the ATG translation start site was set to one. The heterozygous mutation identified with direct double-strand sequencing was further validated with denaturing high-performance liquid chromatography (DHPLC) on a WAVE system (TEAA; Transgenomics, Crewe, England) with the temperature set to 64.6 °C and the start buffer B to 58%. The same DHPLC approach was used to check for the presence of the sequence variant in the DNA of 80 ethnically matched and 96 European control individuals. The role of the variant in splicing was evaluated with in silico analysis using the SplicePort computer program , and DNA sequences of various species encompassing the variant were aligned using options available through Ensembl.
Clinical bilateral anophthalmia was observed in all three affected children of the two families. In family A, both affected children, patients IV:1 and IV:5, had complete absence of eyes, large eyebrows, and a low insertion on the superior eyelid (Figure 2). They had a high arched palate and were normal in height, weight, and head circumference. Both presented normal psychomotor development; general and neurologic examinations were normal for all family members.
As in family A, the affected daughter of family B had large, flaring eyebrows and a low insertion on the superior eyelid (Figure 2). This patient had global development delay, obesity (weight: +7 standard deviation [SD] for age and sex, height: 90th percentile), abnormal head circumference (7th percentile), and a high arched palate.
In patient IV:5 (Figure 3A,C) of family A and the affected daughter (Figure 3B,D) of family B, orbital MRI showed a hypoplastic orbit with present but rudimentary extraocular muscles and a normal lacrimal gland. A cerebral MRI showed agenesis of the optic nerves, optic tracts, and optic chiasma, and a normal pituitary gland. In patient IV:5 (family A), the absence of frontal and sphenoidal sinuses was also noted. In the affected daughter (family B), only the sphenoidal sinus was absent, and significant corticosubcortical atrophy, predominantly in the frontotemporal lobes, was observed with ex vacuo dilation of the ventricles.
Molecular analysis with markers D18S1127 and D18S1147, two markers flanking RAX, showed a homozygous haplotype in the affected individuals of family A suggesting that RAX could be involved in the disease. Sequencing of all the exons and 3′ and 5′ UTR identified a c.543+3A>G mutation (IVS2+3A>G) in the splice donor of the second intron of RAX (Figure 4A). All three affected individuals from both families were homozygous for the mutation while their healthy parents were heterozygous (Figure 4B). The mutation was absent in the healthy sibling (IV:2) in family A (Figure 5). The mutation was not detected in 80 ethnically matched healthy individuals (50 Algerians and 30 Egyptians) and in 96 European controls, and has not been previously reported. No other mutations were detected in OTX2, VAX1, or SOX2.
The third base of the donor site is conserved in various mammals as well as in chickens (Gallus gallus) and fugu fish (Takifugu rubripes). The donor site, however, is no longer conserved in zebrafish and more distantly related species such as Caenorhabditis elegans (Table 2) . In silico analysis of the splice potential donor site of the wild-type sequence of the second intron using the computer program SplicePort gave a score threshold of 1.64346 with a predicted sensitivity of 88.8%, a score usually associated with good splicing potential. This score was reduced to 0.39398 when the IVS2+3A>G variant was evaluated, a score much lower but not indicative of a nonfunctional donor. Haplotype analysis was also performed in family B. The alleles for both microsatellites were different from family A. Based on the analysis of the three individuals available, it seems that a recombination occurred between D18S1127 and D18S1147 in the affected child. No single nucleotide polymorphisms (SNPs) were identified in any of the sequences performed in the three affected individuals.
This is the first report of human bilateral anophthalmia caused by a homozygous mutation of RAX. The identified c.543+3A>G (IVS2+3A>G) splice mutation has not been reported before. Cosegregation analysis in family A confirmed the previously suggested recessive inheritance of RAX mutations in human anophthalmia [10,11]. The mutation affects splicing of the last exon as shown by the strong reduction in splicing potential as determined by SplicePort in silico analysis. This could potentially generate a truncated protein. RAX mutations are rare, and it seems unlikely that two unrelated families would share a similar mutation unless they share a common DNA haplotype. To identify this DNA segment, we performed a haplotype analysis in family B using the same microsatellites flanking RAX. This analysis showed that the three affected individuals do not share the DNA segment defined by the microsatellites used. A smaller common region cannot be excluded and, based on the low frequency of the reported RAX mutations, seems likely. Interestingly, no SNPs were observed in the sequences of the three exons and exon-flanking intronic regions.
Only four reports of RAX mutations in ocular dysgenesis are present in the literature (see Table 3). Voronina et al.  reported a child with unilateral partial anophthalmia (presence of remnants of the globe) due to a compound heterozygous RAX mutation, one missense and one nonsense. Microphthalmia and sclerocornea with persistent fetal vasculature and retinal detachment were present in the other eye. The mother was a healthy carrier of one mutation, and autosomal recessive inheritance was thus suggested. The child was autistic, but no cerebral malformations were observed on the MRI. The second report describes a child with complete bilateral anophthalmia (no remnants of the globe) and normal development and cerebral MRI except for the absence of optic nerves and hypoplastic chiasma. Again, compound heterozygosity was reported but with two nonsense mutations. The parents of the child were not tested for RAX mutations.
Our study confirms the autosomal recessive inheritance of some cases of human anophthalmia, and it seems from the few available cases that two null alleles are necessary to induce such a severe phenotype as anophthalmia. This is suggested by the presence of a healthy mother and a maternal grandmother in the first report  and confirmed by the presence of the four parents of this report who were all heterozygous carriers with normal eye globes. Our report is at odds with a recent report by Gonzales-Rodriguez et al.  in which two patients affected with unilateral anophthalmia and unilateral microphthalmia, respectively, as well as cerebral malformations, were reported to carry a single heterozygous RAX mutation. Neither mutation was found in 400 normal controls suggesting their pathogenicity although the missense mutation carried by the patient with microphthalmia was found to be non-deleterious by the algorithm used by the authors to predict functional impact. The authors concluded that a single RAX allele mutation could cause the presented phenotype. Nevertheless, since we did not perform segregation or functional analysis in this study, the possibility that the second allele carries an undiscovered mutation cannot be ruled out. The fact that the patients reported by Gonzalez-Rodriguez et al.  had a severe phenotype, which includes cerebral malformations, while we report healthy carriers of potentially truncating mutations might support our hypothesis. Moreover, a recent publication on heterozygous RAX mutation associated with ocular dysgenesis described a child with ocular coloboma only and no other associated malformations , suggesting again the mild or normal ocular phenotype that heterozygous RAX mutation carriers can harbor. As a matter of fact, mice heterozygous for the Rax gene mutation develop normal eyes as well .
Although few reports are available, the type of RAX mutation might be determinant for the ocular phenotype. The presence of RAX nonsense mutations is associated with anophthalmia in the reports of Voronina et al. , Lequeux et al. , and Gonzalez-Rodrigues et al.  and this present study. Microphthalmia  and coloboma  are associated with RAX missense mutations. Further studies including functional analyses are required to corroborate this hypothesis.
This is the first report of severe cerebral malformation observed on cerebral MRI associated with homozygous RAX mutations in humans. In addition to the absence of optic nerves, chiasm and optic tracts, which was also reported, albeit in a milder form, by Lequeux et al. , patient IV:5 of family A presented a significant corticosubcortical atrophy with ex vacuo dilation of the ventricles. Rax has been shown to play a major role in brain development in different animal models including mice [8,16]. This observation is thus not surprising and may represent another expression of the severity of the phenotype. Because no cerebral malformation was observed in the case studied by Voronina et al. , these authors attributed a smaller role to RAX in the development of the ventral forebrain than to the morphogenesis of the eye. Our report challenges this conclusion; the type of mutation may instead be incriminated by the presence of cerebral malformations. Further observations, however, are necessary to draw any firm conclusions.
The authors thank Drs. Dalal Shawki, Faten Hussein, Said Gomaa and Fathi ElSahn from the Department of Ophthalmology, University of Alexandria, Egypt, and Dr. Leonidas Zografos from the Jules Gonin Eye Hospital, Lausanne, Switzerland, for their help in making the collaboration between our institutions possible. We thank Isabelle Favre, Tatiana Favez and Natacha Nanchen for technical support.