Article Text


Branchio-oculo-facial syndrome and branchio-otic/branchio-oto-renal syndromes are distinct entities
  1. T Trummer1,
  2. D Müller2,
  3. A Schulze1,
  4. W Vogel1,
  5. W Just1
  1. 1Department of Human Genetics, University of Ulm, Ulm, Germany
  2. 2Klinikum Chemnitz, Säuglingsklinik, Chemnitz, Germany
  1. Correspondence to:
 Dr W Just, Abt Humangenetik, Universitätsklinikum Ulm, Albert-Einstein-Allee 11, 89081 Ulm, Germany,

Statistics from

Branchio-oculo-facial syndrome (BOF, MIM 1136201) is a rare autosomal dominant disorder. The symptoms of this disorder include bilateral postauricular cervical branchial sinus defects with haemangiomatous, scarred skin, cleft lip with or without cleft palate, pseudocleft of the upper lip, nasolacrimal duct obstruction, low set ears with posterior rotation, a malformed, asymmetrical nose with a broad bridge and flattened tip, and, occasionally, prematurely grey hair. Father to son transmission of this disorder has been observed,2 which indicate autosomal dominant inheritance. Another disorder with hearing loss resulting from bilateral branchial cleft fistulas is branchio-oto-renal syndrome (BOR, MIM 113650). Common features of both syndrome are summarised in table 1. Some characteristics of both BOR and BOF syndromes have been reported in a father (BOF) and his son (BOR), but the constant features of BOF syndrome were not observed in either of them. This observation led to the conclusion that BOF and BOR might be allelic variants of the same gene.3 It was suggested that, in both syndromes, penetrance and expression could be variable, and it was concluded that BOF and BOR syndromes are the variable results of mutations in the same autosomal gene.3 However, it was pointed out later that both subjects in fact should be considered as BOF syndrome rather than BOF and BOR syndrome, and that these syndromes are distinct entities and may not be allelic.4 Another related disorder is branchio-otic syndrome (BO, MIM 602588), which comprises branchial fistulas, preauricular pits, and hearing impairment, but lacks renal anomalies (table 1).

Table 1

Branchiogenic disorders

The first candidate gene for BOR has been mapped. This gene, EYA1 (“eyes absent-like”, a human homologue of the Drosophila eyes absent gene), was found by positional cloning5 and maps to chromosome 8q13.3. Mutations in EYA1 have been described,6–8 which made it a candidate gene for BOR syndrome. The authors of the first report8 concluded that BO and BOR syndromes are allelic. The hunt for a candidate gene in BOF syndrome was more difficult, because only a few familial cases exist9 that could be studied. Since an allelic variant of BOF and BOR syndromes was not dispelled conclusively, several independent attempts have been undertaken to study the EYA1 gene region as a candidate gene region for BOF syndrome. By sequence analysis, no mutations were found in the EYA1 gene in five BOF syndrome patients.7 This suggests once more that BOR syndrome might not be allelic to BOF syndrome. EYA1 is a member of a gene family comprising at least four genes (EYA1-EYA4). EYA1 is expressed during embryogenesis in the branchial arches and the somites and during limb development in connective tissue precursors.10 At the tailbud stage of zebrafish, its expression is confined to cranial placodal precursors and, thereafter, to the anterior pituitary, olfactory, and otic placodes.11 The expression of the other members of the EYA gene family, EYA2-3, is similar to the EYA1 expression pattern.10–12 The EYA4 pattern, however, is confined to the dermamyotome and the limb, and expression was not found in the branchial arches.13 The expression patterns in early embryogenesis together with the developmental defects in BOF syndrome prompted a segregation analysis for these four genes in a large pedigree with BOF syndrome, but no cosegregation of the disorder with genetic markers was found.14 The latter study excluded the EYA genes as candidates for BOF syndrome.

Recently, a second gene locus (BOR2) for a BOR syndrome-like phenotype was mapped to human chromosome 1q31.3-q32.1.15 Linkage between the BOR syndrome related disorder branchio-otic syndrome (BO) and marker D1S2757 was observed with a maximum lod score of 4.81 at a recombination fraction of zero.16 The variability of the clinical phenotype of BOF syndrome and the overlapping symptoms with BOR or BO syndromes prompted us to perform a segregation analysis in this second candidate gene region of BO/BOR syndrome in order to verify whether the clinical differences reflect allelic variants of the same gene. In the present study, we show that BOF syndrome is not allelic to BOR/BO syndromes at this locus. This, taken together with previous reports, is a second proof based on genetic studies that BOR and BOF syndromes are distinctive entities. Our findings firmly support former hypotheses on the distinctiveness of these syndromes, which were based solely on the clinical phenotype.


The analyses were performed on DNA of patients with BOF syndrome from one family, which represents the largest published BOF pedigree with five affected and two unaffected members (fig 1). The family studied here was described earlier in a review including photographs of the patients with the characteristic features of BOF syndrome9 (ID 10-14). DNA was extracted from peripheral blood lymphocytes by standard techniques.17 Screening for mutations in the candidate gene region was impossible, because the gene itself is unknown.16 Nevertheless, a segregation analysis with seven markers in an interval of 34 Mb is sensitive enough to exclude cosegregation of a putative candidate gene from this region with the disorder. For this segregation analysis, we used seven microsatellite markers from human chromosome 1q, the candidate region for BO syndrome. The markers span an interval of approximately 34 Mb and their map positions are (from proximal to distal) D1S2815 (204.84 Mb) - D1S461 (209.40 Mb) - D1S2757 (213.2 Mb) - D1S2640 (217.63-217.80 Mb) - D1S2622 (218.20 Mb) - D1S2668 (235.82-235.95 Mb) - D1S249 (238.72 Mb). All given map positions were extracted from NCBI's Entrez Genome Database (URL: The markers include D1S2757 where the highest lod score for BO syndrome has been reported.16 All markers contain dinucleotide repeats, and the maximum heterozygosity of the markers, as published in the database, was always greater than 0.65. Primer sequence information was picked from the GDB database (URL: The DNA of the family members was PCR amplified by a standard protocol with 2.5 mmol/l MgCl2 for 30 cycles comprising three 30 second steps at 95°C, 51°C, and 72°C, respectively, in a “Robocycler Gradient 96” (Stratagene, Amsterdam, The Netherlands). The PCR products were separated on a denaturing 6% polyacrylamide gel in TBE buffer. The DNA was transferred from the gel to a nylon membrane and hybridised with a 32P-labelled oligonucleotide (CA)11. The results from the allele size calculation of the markers of the family were imported into Cyrillic 2.1.3™ pedigree analysis software (FamilyGenetix, Oxford, UK). The preliminary haplotypes were re-evaluated with SimWalk18 to calculate the likelihood of crossovers. The resulting haplotypes were re-entered into Cyrillic for the delineation of the pedigree.

Figure 1

Patient II.3 of the pedigree in fig 2. The fistula of the cervical branchial sinus defect can be identified clearly in addition to the malformation of the low set ear of the newborn girl.


Our results clearly show that BOF syndrome does not cosegregate with markers from the studied region of chromosome 1q (fig 2). Subject III.1 inherited the chromosome with the assumed disease associated markers from her affected paternal grandmother (I.1), whereas III.2 inherited her paternal copy of chromosome 1 from the healthy paternal grandfather (I.2). This information from the haplotypes of grandchildren III.1 and III.2 definitely excludes this region as a BOF syndrome candidate region. The children, II.2 and II.3, inherited the same allele from their mother for marker D1S2815. From the other grandparental chromosome of their mother (these grandparents are not included in the figure), they share marker alleles D1S2640 and D1S2622, the latter with a 50% probability for II.2. Taken together, this clearly shows that BOF syndrome cannot be associated with genes from this region. As long as BOF syndrome is considered to be an autosomal dominantly inherited disorder, which is also evident from the pedigree, the alleles of the markers of the unaffected grandfather I.2 may be disregarded.

Figure 2

Pedigree of the family with BOF syndrome. The haplotypes around the BOR2 locus on chromosome 1q31.3-q32.1 are shown in a 34 Mb interval. The microsatellite markers are arranged (from top to bottom) from proximal to distal 1q. In the studied interval, the affected subjects II.2 and II.3 have inherited an identical haplotype from their healthy father (I.2) but only partially identical (D1S2640 and D1S2622) haplotypes from their affected mother (I.1), which might point to a candidate region. The children (III.1 and III.2) of II.2, however, only have alleles of marker D1S249 from their unaffected mother (II.1) in common, but they do not share any paternal (II.2) alleles. Therefore, the chromosome region around the BOR2 locus on chromosome 1q can be excluded as a candidate region for BOF syndrome, and BOF syndrome is not allelic with BOR syndrome.


Although the phenotypes of the members of the family from our study show intrafamilial variability, their symptoms are all compatible with the diagnosis of BOF syndrome. The phenotypes of patients with BOF syndrome and BOR or BO syndromes have common symptoms, which not only justified the search for mutations in EYA1,7 but also the present study. In addition to previous reports on mutations in the EYA1-4 genes of BOR and BOF syndrome patients,6,7,14 this is the second report, based on genetic analyses, which provides evidence that BOF syndrome and BOR or BO syndromes are phenotypically related but genetically completely distinct disorders which do not represent allelic variants. Based on the syndrome specific symptoms, it has previously been suggested that BOF and BOR/BO syndromes are distinct entities. Now, we provide additionally strong genetic evidence for this assumption with the exclusion of another candidate gene region from chromosome 1q. Unfortunately, the number of affected subjects and families is too small9 to allow for a genome wide linkage analysis.

Obviously, BOF, BOR, and BO syndromes are caused by mutations in different genes. The pathogenetic mechanisms for overlapping and syndrome specific symptoms are unknown, but one might assume that the gene in interval BOR2 on chromosome 1q interacts with EYA1, leading to a cascade of gene activities that regulate proper differentiation of the branchial arches. EYA1 is considered to represent a transcriptional activator, which might interact with downstream genes. Mutations in another transcription factor have been reported recently for the BOR-like Townes-Brocks syndrome (TB syndrome, MIM 10748019,20). This candidate gene SALL1 on chromosome 16q12.1 is the human homologue of Drosophila spalt (sal) which is also a developmental regulator. Therefore, it may be assumed that BOF syndrome might also be caused by mutations in a transcription factor-like protein. A recent report shows expression of SALL1 in the developing brain and the limbs, but also in the lens.21 Expression of EYA1 in the developing eye was excluded by RNA in situ hybridisations in mouse embryos.5 Expression in the eye might be a hint to a candidate gene, because BOF patients occasionally also have anomalies of the eye like coloboma, strabismus, or microphthalmia.


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.