Genetic heterogeneity in patients with enlarged vestibular aqueduct and Pendred syndrome

Abstract

Background: Pathogenic variants in the SLC26A4 gene, encoding for Cl^-/HCO₃^- and I^- anion transporter pendrin, are associated with non-syndromic hearing loss with enlarged vestibular aqueduct (NSEVA) and Pendred syndrome (PDS). In the Caucasian population, up to 75% of patients fail to identify a genetic cause through biallelic mutations in the SLC26A4 gene. The CEVA haplotype could therefore play an important role in the diagnostics of NSEVA. The aim of the study was to determine the genetic etiology of hearing loss with EVA or with fully developed PDS in 37 probands and the functional characterization of novel variants identified in the SLC26A4 gene.

Methods: To determine the genetic etiology, Sanger sequencing, WES and KASP genotyping assay were used. Functional characterization of SLC26A4 variants c.140G>A (p.R47Q), c.415G>A (p.G139R), c.441G>A (p.M147I), c.481T>A (p.F161I), c.1589A>C (p.Y530S) and c.2260del (p.D754Ifs*5) involved determination of iodide influx, total and plasma membrane pendrin expression level and subcellular localization of pendrin by confocal imaging. The nanopore sequencing of nasopharyngeal swab samples was performed to confirm the pathogenic effect of potential splice site variant c.415G>A.

Results: Biallelic variants in the SLC26A4 gene (M2 genotype) were identified in ten probands and a complete CEVA haplotype was confirmed in three probands harbouring SLC26A4 monoallelic variants (M1 genotype). Fifteen variants in the SLC26A4 gene were identified in total, three of which are novel. The functional characterization of the novel variants and variants which were not yet functionally characterized confirmed the pathogenic potential of five out of six tested variants (p.G139R, p.M147I, p.Y530S, p.D754Ifs*5, and p.F161I). Analysis of nasopharyngeal swab samples confirmed exon 4 skipping due to novel variant SLC26A4:c.415G>A. Probands with biallelic SLC26A4 variants had significantly larger thyroid volume per m² of body surface area than subjects with monoallelic SLC26A4 variants and the CEVA haplotype.

Conclusions: The genetic aetiology was determined in 13 out of 37 probands (35%), seven manifested with PDS and six with NSEVA. The present study highlights the importance of functional testing to confirm the pathogenicity of SLC26A4 variants and the phenotype-genotype correlation in SLC26A4-related disorders.

Keywords: SLC26A4; CEVA haplotype; Enlarged vestibular aqueduct; Goiter; Hearing loss.

Fig. 1

Segregation analysis of the variants in the genes SLC26A4 and KCNJ10 and incomplete CEVA haplotype. A Pedigree analysis of proband D92 illustrates the inheritance of variants c.1000+1G>A (p.?) and c.481T>A (p.F161I) in addition to a CEVA haplotype consisting of nine SNPs. The presence of an incomplete homozygous CEVA haplotype alongside a heterozygous hypofunctional p.F161I variant in the unaffected mother (D1984) raises questions about a causal link between the incomplete CEVA haplotype and NSEVA. B Pedigree and pure tone audiometry data for proband D619 and her mother D620 are shown. Both share the variant c.2260del (p.D754Ifs*5) in SLC26A4 and c.301C>A in KCNJ10, with the mother presenting unilateral hearing loss and ipsilateral cochlear hypoplasia without EVA. Proband D619 carries a partial CEVA haplotype consisting of three SNPs

Fig. 2

Identified variants mapped onto the topological model of the pendrin protein. Variants identified in this study are marked with black circles, indicating the affected amino acid residues. Variants selected for functional characterization are underlined with a red dashed line. The red solid line highlights the amino acids deleted in the in-frame loss p.M103_G139del. The topology of distinct domains and N-glycosylation sites (N167 and N172) in pendrin model was created using the Protter tool (Omasits et al. 2014) and based on Saier et al. ; Zheng et al. ; Dossena et al. ; Bassot et al. ; Rapp et al. ; Kuwabara et al.

Fig. 3

Iodide transport efficiency of wild-type pendrin and six pendrin variants. Pendrin function was determined after co-transfection of HEK 293 Phoenix cells with the wild-type (WT) or pendrin variant and the iodide sensor-enhanced yellow fluorescent protein (EYFP) H148Q;I152L or EYFP H148Q;I152L alone (control). A Representative changes in the intracellular fluorescence intensity normalized for the average fluorescence measured before injection of the iodide-containing solution, reflecting pendrin transport efficiency. The arrow indicates the addition of iodide to the extracellular solution. B Percentage of intracellular fluorescence decrease over the experimental period (19 s) determined in cells expressing the WT pendrin or the selected pendrin variants and in control cells. n = 36 measurements from six independent biological replicates, **** p < 0.0001, *** p < 0.001, ns: not statistically significant compared to WT; #### p < 0.0001, # p < 0.05 compared to control, one-way ANOVA with Bonferroni’s multiple comparison post-test

Fig. 4

Subcellular localization of wild-type pendrin and five pendrin variants. Co-localization of pendrin protein with the endoplasmic reticulum (ER) and plasma membrane (PM) was determined in living HeLa cells 72 h after transfection with wild-type (WT) or indicated pendrin variant with EYFP fused to the C-terminus and staining with the ERTracker^™ Red or CellMask^™ Deep Red plasma membrane stain, respectively. A Fluorescent signal of wild-type (WT) or mutant SLC26A4-EYFP (top left), ER (top right), corresponding merged image (bottom left) and scatter plot (bottom right). Scale bar: 25 μm. B Pearson’s correlation coefficient indicated the co-localization of WT pendrin and the studied pendrin variants with the ER. C Fluorescent signal of WT or mutant SLC26A4-EYFP (top left), PM (top right), the corresponding merged image (bottom left) and scatter plot (bottom right). Scale bar: 25 μm. D Pearson’s correlation coefficient indicated the co-localization of WT pendrin and the studied pendrin variants with the PM. **** p < 0.0001, ** p < 0.01 * p < 0.05, ns.: not statistically significant compared to WT, one-way ANOVA with Dunnett’s multiple comparison post-test

Fig. 5

Total protein expression levels of wild-type pendrin and five pendrin variants. HeLa cells were transfected with wild-type (WT) pendrin or the selected variants with EYFP fused to the C-terminus and their expression levels were determined after 72 h using confocal imaging. A Images of fixed HeLa cells showing EYFP signal intensity (red, left panels), corresponding to the expression level of pendrin variants and nuclei counterstained with 4’,6-diamidino-2-phenylindole (DAPI, green, right panels). Scale bar: 100 μm. B Expression levels of WT pendrin and selected variants expressed as EYFP fluorescence intensity normalized for the cell density. n = 24 imaging fields from four biological replicates. **** p < 0.0001, ns.: not statistically significant compared to WT, one-way ANOVA with Bonferroni’s multiple comparison post-test

Fig. 6

Expression level of wild-type pendrin and five pendrin variants in the plasma membrane region. HeLa cells were cotransfected with wild-type (WT) pendrin or the selected variants with EYFP fused to the C-terminus, and transfection marker-enhanced fluorescent protein (ECFP) and their expression levels in plasma membrane (PM) region were determined in living HeLa cells, after staining with the CellMask^™ Deep Red plasma membrane stain, 72 h post-transfection. A Representative images of the indicated SLC26A4-EYFP variants (left panels) and ECFP fluorescent signal (right panels). Scale bar: 25 μm. B Fluorescence intensity of WT and pendrin variants in three regions of the single cell’s plasma membrane (shown in A), averaged and normalized for the fluorescence intensity of ECFP in the cytosol of the same cell. n = 12 cells from two biological replicates. **** p <0.0001, ns.: not statistically significant compared to WT, one-way ANOVA with Dunnett’s multiple comparison post-test

Fig. 7

Effect of variant c.415G>A on pre-mRNA splicing in nasopharyngeal swab-derived samples. A Gel electrophoresis of PCR amplicons generated from cDNA transcribed from RNA of nasopharyngeal swabs. The samples were analysed in duplicates (B1 and B2). Results include endogenous TBP and RPL13A controls, along with amplicons spanning exons 2–6 of the SLC26A4 gene. B IGV visualization of nanopore sequencing reads mapped to the SLC26A4 reference transcript NM_000441.2, illustrating exon 4 skipping associated with variant c.415G>A. C Sashimi plot of reads aligned to the chr7:107,661,600–107,675,150 region of the reference genome, generated using ggsashimi. The x-axis shows the genomic region, while the y-axis represents the read count. Proband D1215 exhibits significant exon 4 skipping, while his father (D2349) shows intermediate coverage of exon 4. Sequencing of the control samples confirmed retention of exon 4 in the transcripts and minimal skipping rate. D Section of the top view of the SLC26A4 protein showing WT (left panel) with amino acids colored according to the ConSurf conservation score. The model with deletion of amino acids from M103 to G139 resulting from exon 4 skipping (middle panel) and its overlap with WT (right panel) show missing conserved α helix in the core domain