Background

Branchio-oto-renal (BOR) syndrome (OMIM113650) is a hereditary dominant autosomal disease with a variable spectrum of manifestations [1]. Currently, most physicians follow the clinical diagnostic criteria based on the principles of Chang [2] to diagnose BOR syndrome. Although scientists reported cases with ear, branchial, and kidney anomalies in the early nineteenth century, Melnick [3] and Fraser [4] were the first to describe these phenotypes in detail. The clinical manifestations of BOR syndrome are highly heterogeneous. Patients with BOR syndrome and ear and branchial defects can be easily identified in early childhood, whereas more time is required for identifying kidney defects [5, 6]. Some cases of kidney abnormalities are detected at infancy, whereas others are detected in adulthood, which delays the appropriate time for therapeutic intervention. This indicates that more attention should be paid to the diagnosis of BOR syndrome as early as possible.

In 1992, Smith and Kumar [7] used linkage studies to locate the pathogenic gene on chromosome 8q13.3. In 1997, scientists [8] studied seven patients with BOR syndrome and identified a novel causative gene, known as EYA1 (eyes absent). Chang et al. [2] demonstrated that approximately 40% patients with BOR syndrome harbored EYA1 mutations. EYA1 is a conserved transcriptional co-activator [9] and is highly expressed in the human embryonic kidney, whereas it is scantily expressed in the brain and lung. In adults, EYA1 is strongly expressed in the heart and skeletal muscles, whereas weaker expression has been observed in the liver and brain; no expression has been observed in the eyes and kidney [10]. In addition, EYA1 is subject to the gene dosage effect, which implies that the quantity of the encoded protein decides the development of the branchial arch, ear, and kidney; the gene activity is discernible only if the encoded protein quantity surpasses a certain threshold. This can explain the differences in phenotypes among BOR patients in one family.

Although previous studies have reported many BOR-related mutations [11,12,13,14,15,16], this is the first time that BOR syndrome has been investigated genetically and pathogenic mutations in EYA1 have been detected in a Chinese family with BOR syndrome.

Methods

Subjects

We collected the pedigree information of 17 members of a family over three generations at the First Affiliated Hospital of Zhengzhou University, China (Fig. 1).

Fig. 1
figure 1

Pedigree of a family with branchio-oto-renal (BOR) syndrome

Full size image

Figure 1 shows the relationship and the prevalence situation of this pedigree with BOR syndrome. The four main features of BOR syndrome are identical to the criteria used for distinguishing disease progression (renal abnormalities, branchial fistula, hearing loss, and preauricular pits) [2]. The other family members of II:1 (I:1, I:2, II:4, II:5, II:7, III:3, III:4, III:5, and III:6) were normal.

Sanger test

Venous blood was obtained for screening EYA1 mutations from 10 living members (I:1, I:2, I:3, II:1, II:2, II:4, II:6, III:1, III:2, and III:6) of this family after obtaining informed consent. Genomic DNA was isolated using the AxyPrep blood genomic DNA miniprep kit (Axygen Biosciences, CA, USA) per manufacturer’s instructions. Eighteen primers corresponding to 18 exons of EYA1 (NM_000503) were designed using the online tool Primer3web (http://primer3.ut.ee/) (Additional file 1: Table S1). The sequencing was performed using an ABI BigDye Terminator cycle sequencing kit (Applied Biosystems, CA, USA) on an ABI PRISM 3730 DNA Analyzer (Applied Biosystems), and the data were analyzed using the software CodonCode Aligner and Mutation Surveyor V4.0.5. SIFT, Polyphen2, and MutationTaster [17] were used to predict the pathogenicity of the variant. The mutation site was extensively analyzed using the online tool MutationTaster (http://www.mutationtaster.org/).

Minigene experiment

The genomic DNA of III:1 (control sample) and III:2 (case sample) were PCR amplified with Q5 hot start high-fidelity DNA polymerases (New England Biolabs, MA, USA) for exon 11 of EYA1 (2,051 bp total, 967 bp upstream, and 1,000 bp downstream, according to c.967A). Next, the PCR products were digested with XbaI and XhoI (New England Biolabs) and ligated in carrier pCMV-Tag-2B (Agilent Technologies) to generate pCMV-con (c.967A) and pCMV-case (only c.967 T was picked). The two plasmids were transformed in 293 T cells using Lipofectamine 2000 (Invitrogen Life Technologies, Carlsbad, CA, USA) following the manufacturer’s instructions. After 12 h, the medium was changed to Dulbecco’s modified Eagle medium plus 10% fetal bovine serum (Gibco) and incubated for another 24 h till the total RNA levels reached the maxima. Then, the two plates of cells (pCMV-con and pCMV-case) were treated with Trizol, and RNA samples were obtained. We sequenced the two cDNAs synthesized from the extracted RNA using the PrimeScript™ RT reagent kit (Takara, Osaka, Japan).

Western blot analysis

The cDNA of EYA1 was inserted into pRK7-N-Flag between XbaI and EcoRI to construct the vector pRK7-conl. c.967A was mutated to c.967 T in the pRK7-conl vector to obtain pRK7-case using the Takara MutanBEST kit (Takara). 293 T cells, which were serially transfected with pRK7-conl and pRK7-case, were washed in phosphate-buffered saline (PBS), and the supernatant was completely discarded. Next, two plates of cells were resuspended in 1X sodium dodecyl sulfate (SDS) buffer, followed by incubation at 99°C for 10 min. The scattered resuspension solution was collected. The proteins were separated on 10% SDS polyacrylamide gels at 80 V for 20 min and then at 130 V for 1 h. The proteins were semidry-blotted onto polyvinylidene fluoride membranes at 320 mA for 1 h. After blocking with 5% milk in PBS-Tween 20 (PBST) for 1 h, the membrane was incubated with mouse anti-FLAG monoclonal antibodies (Invitrogen) overnight at 4°C. The membrane was washed thrice with PBST for 30 min, incubated for 1 h with rabbit anti-mouse IgA/HRP (Invitrogen), and washed thrice for another 30 min. The membrane was incubated in enhanced chemiluminescence (ECL) Plus for 2 min, and the emitted light was detected using a CCD camera.

Results

Clinical observations

As shown in Fig. 1, III:2 is the proband. Subject II:2 and her children (III:1, III:2) had bilateral preauricular pits. III:1 presented with bilateral branchial cleft fistulae and microtia (Fig. 2). Pure tone audiometry of III:1 showed bilateral conductive hearing loss. Bilateral middle ear malformation and inner ear hypoplasia were observed in the computed tomography (CT) scan (Fig. 3).

Fig. 2
figure 2

Images of both ears with microtia, preauricular pits, and branchial cleft fistulae in subject III:5. a and b Small arrows indicate preauricular pits, and large arrows indicate microtia. c Small arrows indicate branchial cleft fistulae, and large arrows indicate microtia

Full size image
Fig. 3
figure 3

Axial computed tomographic images of the temporal bone from subject III:5. The small arrow indicates a malformed middle ear, including anteriorly displaced abnormal ossicles and hypoplastic mastoid cells. The middle arrow indicates malformed cochlea. The large arrow indicates microtia

Full size image

Subject III:2, the 5-year-old proposita, underwent a hearing test when she was 3 years old. Auditory brainstem response (ABR) testing revealed otitis media in both ears. The threshold of bone-conducted ABR was 20 dB nHL for both sides, while the threshold of air-conducted ABR was 55 dB nHL on the left side and 60 dB nHL on the right side. She had both microtia and facial hypoplasia on the right side. She also had palatoschisis and bilateral branchial cleft fistulae. Magnetic resonance imaging (MRI) showed bilateral otitis media and bilateral posterior semicircular canal and cochlear dysplasia. MRI also demonstrated thinner bilateral facial nerves than normal.

Subject II:6 had bilateral mixed hearing loss and preauricular pits. He had branchial cleft fistula only on the left side. Hydronephrosis and abnormal renal function had been detected several years ago. However, these symptoms were not present at the recent clinical evaluation. Renal ultrasonography showed bilateral mild diffuse echo change and slightly separated collecting system on the right side. CT showed bilateral inner ear hypoplasia and otitis media on the right side. His 3-year-old son (III:6) was negative for these findings by ABR, CT, and renal ultrasonography; however, his speech was not normal.

Genetic observations

All 18 exons of EYA1 and their flanking intron sequences were analyzed for all the living members in this family. Direct sequencing of the EYA1 gene revealed that all the affected patients (II:2, III:6, III:1, and III:2) in this family carried a novel heterozygous nonsense variation (c.967A > T according to the GenBank transcript ID: NM_000503; p: Arg323X) in exon 11 (Fig. 4). The mutation was located at the beginning of exon 11 of EYA1. Meanwhile, none of the unaffected family members (I:1, I:2, I:3, II:1, II:4, or III:6) carried the mutation in EYA1. This mutation was not observed in any current database, including the HapMap, ExAC, and 1000 Genomes projects. Thus, the mutation is unlikely to represent genetic polymorphism.

Fig. 4
figure 4

EYA1 gene test in this family revealed a novel heterozygous nonsense variation (c.967A > T; p: Arg323X) in exon 11 in members with various physical anomalies (II:2, III:6, III:1, and III:2) (arrow). The normal family members harbored no mutations in EYA1

Full size image

Minigene results

Agarose gel electrophoresis results showed the wild-type and mutant gene to have a band of the same size (Additional file 2: Figure S1). cDNA sequence analysis confirmed the transcripts as wild-type, consistent with the reference sequence (Fig. 5). Although in silico analysis suggested the possibility of aberrant splicing being caused by c.967A > T, we did not detect any abnormal transcripts. The results showed that the mutation did not affect splicing, which caused a truncated protein to be the genetic cause of BOR syndrome in this family.

Fig. 5
figure 5

Results of the minigene experiment. Minigene experiments showed no change in RNA level according to the point change

Full size image

Western blot analysis

Western blotting was performed to visualize the expressed proteins with a molecular weight of ~ 65 kDa for wild-type EYA1. The results showed that both wild-type and mutant proteins were expressed, but the mutant protein was significantly shorter, about half the size of the wild-type protein. This further proves that c.967A > T may lead to the premature termination of the protein (Fig. 6).

Fig. 6
figure 6

Western blotting shows that the protein encoded by the mutated plasmid was truncated compared to the wild type

Full size image

Discussion

BOR syndrome is one of the most common autosomal dominant syndromes, characterized by five major phenotypes, namely, hearing loss (about 98%), preauricular pits (about 83%), branchial fistulae (about 68%), renal anomalies (about 38%), and external ear abnormalities (about 31.5%), based on studies on 40 families (Table 1) [2]. Some patients with branchial cleft, hearing loss, and preauricular pits have been found with branchio-otic (BO) syndrome [18], whereas others with hearing loss, branchial cleft, and obvious ureteral deformity are evaluated with branchio-oto-ureteral (BOU) syndrome [19]. It is well known that BO and BOU syndromes share the same pathogenic gene: EYA1 [8, 20,21,22].

Table 1 Criteria for BOR syndrome
Full size table

EYA1 is a member of the EYA family, and other members include EYA2, EYA3, and EYA4 [23]. The online website http://www.hgmd.cf.ac.uk/ac/index.php records 164 mutations related to BOR syndrome (last accessed, April 2016), of which 152 mutations are in EYA1, including missense, nonsense, and frameshift mutations [24,25,26].

The clinical phenotypes of BOR syndrome showed high heterogeneity [27]. Phenotypic variations are common for affected lineages because of the incomplete penetrance and variable expressivity within or between families. The affected members of this family also showed different phenotypes. For example, II:6 presented with all symptoms of BOR syndrome. III:1 and III:2 had preauricular pits, hearing loss, and branchial fistulae, whereas II:2 had only preauricular pits and hearing loss, indicating that II:2 strictly showed BO syndrome characteristics.

Genetic testing is increasingly becoming an important tool for clinical diagnosis. Genetic testing for this family revealed that all patients had a novel variation (c.967A > T, p: Arg323X) in exon 11 of EYA1. The variation produced a premature termination codon, which caused premature termination of translation. The online tool MutationTaster predicted that an alternative splicing site was located at the beginning of exon 11, which could possibly be altered by the mutation. Based on this hypothesis, minigene analysis was used to investigate whether this mutation affected splicing. However, no difference was observed between the mutant and the control. Hence, we speculated that this change produced only a termination codon.

Xu [28] reported an Eya1-knockout mouse model with hearing and renal defects and aberrant organ development. In addition, the morphogenesis of some parts, such as the thymus, parathyroid, and thyroid, was affected [29]. EYA1 is a key gene for mammalian organogenesis, mutations in which result in multiple organ malformation. Similar to other EYA family members, EYA1 possesses a highly conservative 271-amino acid C-terminal EYA domain and a divergent N-terminal transactivation domain for protein–protein interactions [30]. Although EYA1 is a transcriptional activator, it did not bind directly to DNA in in vitro experiments. Co-factors of EYA1 are members of the SIX family of proteins. All SIX proteins possess two parts: one is the highly conserved SIX domain (SD) with DNA-binding ability and the other is an N-terminal domain that interacts with other proteins, including EYA. Several studies have shown that most of the SIX members and EYA1 are expressed during the development process. Interestingly, SIX1, SIX5, and EYA1 are associated with BOR syndrome. Among these, EYA1 mutations occur in approximately 41% of patients with BOS1 or BOR1 syndrome, and SIX1 mutations affect 3.5~ 4.5% of patients with BOS3 or BOR3 syndrome, whereas SIX5 has been mentioned in only one report. Furthermore, the association of the SIX5 mutation with BOR syndrome has been recently disputed [31], as Six−/− mice were found to develop only cataract. SIX1 and EYA1 are expressed in the growing basal plates, which include the ear placodes and the developing kidney [32], and they regulate the expression of other genes that are critical for the development of sense organs [33]. Loss of function of these two genes in other species causes placodal defects, alterations in the gene expression profile, cell apoptosis, reduction in cell proliferation in several placodes, etc. [28, 29].

Since the role of EYA1 in development is crucial, it is related to four diseases as a pathogenic gene: otofaciocervical syndrome, anterior segment anomaly, BO syndrome, and BOR syndrome. Because the known pathogenic mutations of EYA1 are loss-of-function mutations, haploinsufficiency might be the main pathological mechanism (Fig. 7).

Fig. 7
figure 7

Mutation in EYA1 exon 11 generates a stop codon. The haploinsufficiency of normal EYA1 disrupts SIX1 binding, which is necessary for the activation of target genes

Full size image

Owing to the strong heterogeneity of clinical phenotypes, it is sometimes difficult for BOR syndrome to be diagnosed only based on clinical phenotypes, necessitating genetic testing for its early diagnosis. Early diagnosis would be beneficial for patients with BOR syndrome, as they could be sensitive to lifestyle changes and monitoring of kidney function later in life.

In this study, we identified a novel mutation in EYA1; however, we have not investigated the molecular mechanisms via which this mutation affects the pathogenesis and prognosis of the disease, which warrants further investigation. In addition, a limitation of this study is that the complete medical details of all the family members are not available.

Conclusion

In summary, we report a novel nonsense mutation in EYA1 as a causative mutation for BOR syndrome. We highlight that combining molecular tests with the analysis of clinical phenotypes would contribute to the timely diagnosis and treatment of BOR syndrome.