Abstract
Purpose: :
To investigate mRNA splicing for four missense mutations (c.1450G>C/p.A484P; c.3625A>G/p.T1209A; c.4520G>A/p.R1507Q and c.5237G>A/p.R1746Q) that have been described in USH1 patients in homozygous state, thereby constituting an exception from the genotype-phenotype correlation for CDH23 mutations. Truncating mutations usually result in Usher syndrome type 1D (USH1D), whereas missense mutations affecting highly conserved motifs of the CDH23 protein cause non-syndromic deafness (DFNB12).
Methods: :
Conservation across species (human, mouse, chicken and zebrafish) and between the 27 CDH23 ectodomains (EC) was investigated by applying PIMA 1.4 and Boxshade 3.21 to the four missense mutations and 27 non-synonymous CDH23 SNPs previously reported. Prediction of splicing was conducted using NNSplice and the ESE finder. A minigene assay was applied for analysing mRNA splicing in COS-7 cells. Moreover, to assess the long-term clinical impact of two in-frame mutations (p.M1281del and c.7362+5G>A) in a then 15-year-old patient previously reported by us, we re-examined him at age 21.
Results: :
p.A484, p.T1209, p.R1507 and p.R1746 all show evolutionary conservation. While p.A484 shows conservation between human CDH23 EC domains, p.R1507 does not; p.T1209 and p.R1746 reside in linker regions. Of the previously reported 27 non-synonymous SNPs, 14 (52%) were evolutionarily conserved. Of the 25 SNPs that were located in cadherin repeats, five (20%) showed strong conservation between the different EC domains. While in silico analysis suggested impairment of splicing in all four cases, we found aberrant splicing for only one mutation, p.R1746Q. However, splicing was normal in case of p.A484P, p.T1209A and p.R1507Q. The degree of disease progression does not deviate perceivably from what would be expected in a typical USH1 patient.
Conclusions: :
Neither comparison with paralogues, nor the alignment of the 27 ECs or exonic enhancer prediction with ESE finder are helpful for the assessment of human CDH23 missense variants. Programs such as SpliceView and NNSplice - although potentially helpful - can not replace experimental verification. p.A484P, p.T1209A and p.R1507Q could interfere with functions of both, the auditory and the visual system. Alternatively, they could represent rare non-pathogenic polymorphisms.
Keywords: gene/expression • genetics • mutations