The multi-faceted nature of 15 CFTR exonic variations: Impact on their functional classification and perspectives for therapy

BACKGROUND

The majority of variants of unknown clinical significance (VUCS) in the CFTR gene are missense variants. While change on the CFTR protein structure or function is often suspected, impact on splicing may be neglected. Such undetected splicing default of variants may complicate the interpretation of genetic analyses and the use of an appropriate pharmacotherapy.

METHODS

We selected 15 variants suspected to impact CFTR splicing after in silico predictions on 319 missense variants (214 VUCS), reported in the CFTR-France database. Six specialized laboratories assessed the impact of nucleotide substitutions on splicing (minigenes), mRNA expression levels (quantitative PCR), synthesis and maturation (western blot), cellular localization (immunofluorescence) and channel function (patch clamp) of the CFTR protein. We also studied maturation and function of the truncated protein, consecutive to in-frame aberrant splicing, on additional plasmid constructs.

RESULTS

Six of the 15 variants had a major impact on CFTR splicing by in-frame (n = 3) or out-of-frame (n = 3) exon skipping. We reclassified variants into: splicing variants; variants causing a splicing defect and the impairment of CFTR folding and/or function related to the amino acid substitution; deleterious missense variants that impair CFTR folding and/or function; and variants with no consequence on the different processes tested.

CONCLUSION

The 15 variants have been reclassified by our comprehensive approach of in vitro experiments that should be used to properly interpret very rare exonic variants of the CFTR gene. Targeted therapies may thus be adapted to the molecular defects regarding the results of laboratory experiments.

Functional characterization and phenotypic spectrum of three recurrent disease-causing deep intronic variants of the CFTR gene

BACKGROUND

The CFTR genotype remains incomplete in 1% of Cystic Fibrosis (CF) cases, because only one or no disease-causing variants is detected after extended analysis. This fraction is probably higher in CFTR-Related Disorders (CFTR-RD). Deep-intronic CFTR variants are putative candidates to fill this gap. However, the recurrence, phenotypic spectrum and full molecular characterization of newly reported variants are unknown.

METHODS

Minigenes and analysis of CFTR transcripts in nasal epithelial cells were used to determine the impact on CFTR splicing of intronic variants that we previously identified by next generation sequencing of the whole CFTR locus. Phenotypic data were collected in 19 patients with CF and CFTR-RD, in whom one of the deep intronic variants has been detected.

RESULTS

Three deep-intronic variants promoted the inclusion of pseudo-exons (PE) in the CFTR transcript, hindering the synthesis of a functional protein. The c.2989-313A > T variant, detected in four patients with CF or CFTR-RD from three different families, led to the inclusion of a 118 bp PE. The c.3469-1304C > G variant promoted the inclusion of a 214 bp-PE and was identified in five patients with CF from four families. Haplotype analysis confirmed that this variant was associated with one CF chromosome of African origin. The most represented variant in our cohort was the c.3874-4522A > G, detected in 10 patients with various phenotypes, from male infertility to CF with pancreatic insufficiency.

CONCLUSION

These three deep intronic CFTR variants are associated with a large phenotypic spectrum, including typical CF. They should be included in CF diagnostic testing and carrier screening strategies.

A new large CFTR rearrangement illustrates the importance of searching for complex alleles

The p.Val754Met variant, described in 1996 in a CF patient, has been considered a CF mutation. However, biochemical aspects, results of functional studies and, finally, the identification of a complex deletion removing exons 3 to 10 and 14b to 16 in cis of p.Val754Met in a CF patient, argue against a strong deleterious effect. An inventory through the French CF network of patients carrying p.Val754Met led to the registration of seven patients (CF: n=4; idiopathic chronic pancreatitis: n=3) and six healthy individuals, all heterozygous for the variation. Extensive CFTR gene analysis was carried out, including the search for large rearrangements and other possible mutations. The complex deletion, whose breakpoints are described here, was found only in the four CF patients, in association with the same haplotype. This data, added to the fact that the p.[Phe508del]+[Val754Met] genotype was found in a healthy individual, bring further arguments against the association of p.Val754Met with CF. We thus suggest looking for a possible complex allele whenever p.Val754Met is detected and considering it neutral regarding genetic counseling when found in isolation.

Spectrum of CFTR mutations in cystic fibrosis and in congenital absence of the vas deferens in France

We have collated the results of cystic fibrosis (CF) mutation analysis conducted in 19 laboratories in France. We have analyzed 7, 420 CF alleles, demonstrating a total of 310 different mutations including 24 not reported previously, accounting for 93.56% of CF genes. The most common were F508del (67.18%; range 61-80), G542X (2.86%; range 1-6.7%), N1303K (2.10%; range 0.75-4.6%), and 1717-1G>A (1.31%; range 0-2.8%). Only 11 mutations had relative frequencies >0. 4%, 140 mutations were found on a small number of CF alleles (from 29 to two), and 154 were unique. These data show a clear geographical and/or ethnic variation in the distribution of the most common CF mutations. This spectrum of CF mutations, the largest ever reported in one country, has generated 481 different genotypes. We also investigated a cohort of 800 French men with congenital bilateral absence of the vas deferens (CBAVD) and identified a total of 137 different CFTR mutations. Screening for the most common CF defects in addition to assessment for IVS8-5T allowed us to detect two mutations in 47.63% and one in 24.63% of CBAVD patients. In a subset of 327 CBAVD men who were more extensively investigated through the scanning of coding/flanking sequences, 516 of 654 (78. 90%) alleles were identified, with 15.90% and 70.95% of patients carrying one or two mutations, respectively, and only 13.15% without any detectable CFTR abnormality. The distribution of genotypes, classified according to the expected effect of their mutations on CFTR protein, clearly differed between both populations. CF patients had two severe mutations (87.77%) or one severe and one mild/variable mutation (11.33%), whereas CBAVD men had either a severe and a mild/variable (87.89%) or two mild/variable (11.57%) mutations.