Article Text


The gene for branchio-oculo-facial syndrome does not colocalise to theEYA1-4 genes
  1. * Department of Medical Genetics, University of Ulm, Albert-Einstein Allee-11, 89069 Ulm, Germany
  2. Doctorado en Genética Humana, Universidad de Guadalajara, Guadalajara, México
  3. Klinikum Chemnitz, Säuglingsklinik, Chemnitz, Germany
  1. Professor Vogel,walther.vogel{at}

Statistics from

Editor—The branchio-oculo-facial syndrome (BOFS) is characterised by a branchial cleft sinus or linear skin lesion behind the ear, lacrimal duct obstruction, colobomata of the iris/retina, hypertrophy of the lateral pillars of the philtrum (“pseudocleft”), an asymmetrical nose with a broad tip, and auricular and lip pits. Premature greying of the hair is also observed.1 Inheritance is autosomal dominant (OMIM 113620).2 Several anomalies common to both BOF and BOR (branchio-oto-renal) syndromes have been reported.3 McCool and Weaver4 reported three cases with BOF and unilateral renal agenesis. This anomaly is not frequent in BOFS but is characteristic for patients with BOR, and hence a contiguous gene syndrome or the presence of different mutations within a single gene have been suggested.4 Recently, the BOR gene was identified by positional cloning on chromosome 8q13.3 and mapped between markers D8S1060 and D8S1807.5 The gene was named “eyes absent-like 1” (EYA1), the human homologue of the Drosophila eyes absent gene. It has been postulated that the EYA1 gene and two otherEYA related genes (EYA2 on chromosome 20q13.1 and EYA3 on 1p36) play a role in development.5

Based on the largest published family with BOFS6 and in order to find a candidate gene for BOFS, we studied four flanking markers in the BOR chromosome region (EYA1 gene)7 as well as six markers flanking the EYA2gene,8 four markers at the EYA3gene,9 and four markers close to theEYA4 gene, which has recently been mapped to 6q23.10

The family studied here was described in detail by Linet al 6 (patients 10-14) and includes five affected and two unaffected members. Autosomal inheritance is shown by a father to son transmission. The affected subjects show intrafamilial variability but their symptoms are all compatible with the clinical diagnosis of BOFS.

Genomic DNA was prepared from peripheral blood lymphocytes from all family members using standard procedures. Four microsatellite markers (D8S543, D8S530, D8S279, and D8S286, covering 7 cM) flanking theEYA1 gene on chromosome 8q, six markers flanking the EYA2 gene on chromosome 20q (D20S899, D20S721, D20S911, D20S119, D20S836, and D20S17, covering 8 cM), four microsatellite markers around theEYA3 gene on chromosome 1p (D1S2893, D1S214, D1S244, and D1S228, covering 21 cM), and four microsatellite markers around EYA4 on 6q23 (D6S1656, D6S413, D6S270, and D6S292, covering 4 cM) were analysed. PCR amplification was performed on 50 ng of DNA using fluorescently labelled primers (the markers were chosen from GenBank and the primer sequences were taken from The PCR products were run on 6% denaturing polyacrylamide gels in a fluorescence sequencer and analysed with the AlleleLinksTM program (Amersham Pharmacia Biotech). The results were exported to Cyrillic 2.1TM for pedigree drawing. The segregation of the haplotypes was determined using SimWalk,11, the resulting haplotypes re-entered into Cyrillic, and evaluated for cosegregation with BOFS.

The haplotypes at the four loci were determined by SimWalk11 with high probability. The pedigrees with the haplotypes at the different EYAn loci are shown in fig 1. The haplotypes of the EYA1locus obviously segregate independently of the disease. The children (III.1 and III.2) of II.1 inherited different chromosomes but are both affected. Cosegregation of the BOF syndrome with theEYA2 gene could also be excluded because different haplotypes were passed to the affected subjects in generations II and III (markers D20S721 and D20S911 were not informative and were excluded from fig 1). A similar situation occurs for the haplotypes around the EYA3 locus with different haplotypes segregating from the affected mother (I.2) to the affected children (II.1 and II.3). Another apparent example of exclusion of cosegregation of the disorder with a gene locus is shown for EYA4, where each haplotype of the affected mother (I.2) was passed to the affected children (II.1 and II.3), whereas only one of the haplotypes around theEYA4 locus of the unaffected father (I.1) could be found constantly in all affected children. The latter finding is just by chance and the shared haplotype for the paternal (I.1) chromosome 6 in all affected children does not contribute to the phenotype.

Figure 1

Pedigrees of the family affected with BOF syndrome showing the haplotypes around the EYA1, 2, 3, and 4 genes. Filled symbols represent patients, empty symbols correspond to unaffected subjects. Cosegregation of the disease locus of the BOF syndrome with EYA1 can be excluded because II.1 and II.3 inherited different chromosomes from their mother (I.2). Cosegregation of the disease with the EYA2 gene can be excluded comparing III.1 and III.2. Cosegregation of the disease with the EYA3 gene can be excluded because III.1 and III.2 have inherited the “unaffected” haplotype from their father (II.1). A shared haplotype at the EYA4 locus cannot be found in this pedigree, excluding EAY4 as a BOFS candidate gene too. Markers used (top to bottom): EYA1: D8S543 - (EYA1) - D8S530 - D8S279 - D8S286 7 cM, EYA2: D20S899 - D20S119 - (EYA2) - D20S836 - D20S17 8 cM, EYA3: D1S2893 - D1S214 - (EYA3) - D1S244 - D1S228 21 cM, EYA4: D6S1656 - D6S413 - D6S270 - (EYA4) - D6S292 4 cM.

From these segregation analyses in the largest family with BOF syndrome reported to date, we conclude that BOF and BOR syndrome may not be allelic. We did not find cosegregation of the disease with the markers from the critical region of the BOR syndrome (EYA1) or with the related genesEYA2, EYA3, orEYA4.

The BOF and BOR syndromes were originally postulated to represent one contiguous gene syndrome because of overlapping clinical features.3 4 However, the variable expression of the BOR syndrome especially with respect to renal anomalies resulted in the delineation of a BO (branchio-oto) syndrome. Recently, it was shown that BOR and BO are allelic defects ofEYA1.12 Our patients do not have renal abnormalities typical of BOR syndrome, but share branchial and otological alterations found in BO and BOF syndromes. Intrafamilial variability in BOFS is more indicative of an allelic disorder12 than of a contiguous gene syndrome. This intrafamilial variability also illustrates the difficulty in delineating distinct syndromes based only on isolated cases.

EYA1 is a member of theEYA gene family which at present comprises four genes.10 It has been postulated that all theEYA gene family members may cause developmental defects when mutated.5 10 The three first genes (EYA1-3) are expressed in the ninth week of human development.5 Hence,EYA2 and EYA3could be other candidate genes for BOFS.EYA4, a new gene of theEYA family, was recently identified.10 This gene has been localised to chromosome band 6q23. Unlike the other three EYA genes it is not expressed in the developing eye, but in the early developing mouse embryo it is expressed in the otic vesicle, the branchial arch region, and in the craniofacial mesenchyme above the nasal process and between the eyes. This expression profile caused us to study a possible cosegregation of EYA4 with BOF syndrome, which is characterised by branchial and otological alterations. Obviously, EYA4 is not a candidate gene for BOF syndrome.

From our data, we can conclude that the known genes of theEYA family are not involved in the BOFS and BOFS is not allelic to BOR syndrome. Hitherto unknown genes from this gene family cannot be excluded as candidate genes. Final exclusion of the EYA genes as candidates, however, can be done only when our data are confirmed by studies on other families. We are aware that genetic heterogeneity may lead to the exclusion ofEYA genes in our family, while in other families EYA may be involved in BOFS. However, the rarity of such cases makes it difficult to present an undisputed candidate gene. A genome wide search for other genes using a panel of polymorphic markers will help in the search for the candidate gene for BOF syndrome.


LCC is supported by a DAAD fellowship. We thank Thea Trautmann for skilful technical assistance.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.