M. des Georges

Preimplantation genetic diagnosis for cystic fibrosis: the Montpellier center's 10-year experience

This study provides an overview of 10 years of experience of preimplantation genetic diagnosis (PGD) for cystic fibrosis (CF) in our center. Owing to the high allelic heterogeneity of CF transmembrane conductance regulator (CFTR) mutations in south of France, we have set up a powerful universal test based on haplotyping eight short tandem repeats (STR) markers together with the major mutation p.Phe508del. Of 142 couples requesting PGD for CF, 76 have been so far enrolled in the genetic work‐up, and 53 had 114 PGD cycles performed. Twenty‐nine cycles were canceled upon in vitro fertilization (IVF) treatment because of hyper‐ or hypostimulation. Of the remaining 85 cycles, a total of 493 embryos were biopsied and a genetic diagnosis was obtained in 463 (93.9%), of which 262 (without or with a single CF‐causing mutation) were transferable. Twenty‐eight clinical pregnancies were established, yielding a pregnancy rate per transfer of 30.8% in the group of seven couples with one member affected with CF, and 38.3% in the group of couples whose both members are carriers of a CF‐causing mutation [including six couples with congenital bilateral absence of the vas deferens (CBAVD)]. So far, 25 children were born free of CF and no misdiagnosis was recorded. Our test is applicable to 98% of couples at risk of transmitting CF.

WS17.2 Identification of CF mutations in deep intronic regions: Design of antisense oligonucleotides for a targeted therapeutic approach

Considering that the extensive study of the CFTR gene classically performed in molecular diagnosis does not detect all disease-causing mutations, we previously developed an approach for a complete resequencing of the CFTR locus to search for mutations deeply located in introns (Bonini et al. Genet Med, 2015). After identifying candidate intronic mutations, we now aim to restore full-length CFTR transcripts by using antisense oligonucleotides also named Target Site Blockers (TSB). DNA from 15 patients with only one CF mutation were collected through a national collaborative study. Enrichments of the CFTR locus (250 kb) by long-range PCR or targeted capture were sequenced using 454 GS Junior and MiSeq Illumina platforms. TSB (Exiqon) were specifically designed for the deleterious variants. Out of 200 variations detected, our in-house pipeline pointed out 10 intronic variations, per DNA in average. In silico analysis tools predicted that 13 variants (for the 15 patients) had a deleterious effect on CFTR splicing by promoting inclusion of pseudoexons. TSB were tested whether the splicing defect was confirmed. For instance, TSB used on bronchial and primary nasal cells transfected with minigene constructs, significantly restored the full-length CFTR transcript for the well-known mutation c.1680–886A>G and a new identified c.1680–883A>G. Two new mutations in introns 18 and 23 are being tested. Finally, our massively parallel sequencing strategy lead to the identification of new CF-causing mutations in introns. Our findings also demonstrate the efficiency of antisense oligonucleotides for a targeted therapeutic for cystic fibrosis. Work supported by Vaincre La Mucoviscidose.

3 Evaluation of the Multiplicom CFTR MASTR V2 kit on a Roche 454 Junior NGS platform as a diagnostic tool for a comprehensive analysis of the CFTR gene

Twenty-three years after the discovery of the CFTR gene, more than 1900 anomalies are described in the world, the majority of them being single base-pair substitutions or micro-insertions/deletions. Identification of mutations has important implications for genetic counselling, prenatal diagnosis, cascade screening in families, as well as for understanding the genotype–phenotype relationship. Since a few years, NGS technologies enable us to overcome the classical approaches of whole coding sequence sequencing at single nucleotide resolution. The aim of our study is to compare 3 different strategies with the Ion Torrent technology (Life Technologies, LT): Ampliseq (LT), Haloplex (Agilent) and homemade Long Range PCR (LR). Preliminary results showed that the coverage of the CFTR locus by Ampliseq is limited to 51% and includes 86% of exons with a depth generally consistent between amplicons. Haloplex approach gives more heterogeneous depth and coverage. Finally, LR-PCR showed coverage consistent with the design but differences at depth between each LR amplification. The development of libraries by the two first approaches is rapid (less than 24 h) and technically easy. For LR-PCR, It takes about two days to build a library with long and time-consuming manual steps. Sequencing result for the CFTR gene were validated with wild-type samples and DNA carrying known variants (mutations and polymorphisms) previously identified by DHPLC, HRM and sequencing. In conclusion, the sequencing of the entire CFTR locus by NGS is a tool that can quickly respond to issues such as prenatal diagnostic and the most promising approach seems to be the LR strategy.

WS21.2 Non-invasive prenatal diagnosis (NIPD) of cystic fibrosis by quantitative real time mutant enrichment with 3′-modified oligonucleotides (MEMO) PCR

Objectives: The incidence of Cystic Fibrosis (CF) in the European Caucasian population is about 1 in 2500 live births and about 1 in 25 people are carriers of a CF genetic mutation. These numbers were generally adopted, but there may be marked regional variations. The p.Phe508del mutation is the most common mutation causing CF. In central and northern Europe, this mutation has a frequency of about 70%. But previously published data indicated that in Spain and Portugal the frequency of this mutation was much lower, around 50%. To estimate the incidence of CF in Portugal we have determined with high accuracy the frequency of the p.Phe508del in our CF patients and calculate the incidence of the disease by determining the incidence of CF carriers analysing consecutive new-borns. Methods: A group of 509 CF patients, with extensive or complete CFTR gene analysis (by DGGE and/or sequencing), was clinically revaluated and part of the diagnostics were excluded. The frequency of the p.Phe508del mutation was determined in the confirmed CF cases as 66.4%±3.2% (95% CI). Then, we have analysed 1603 Guthrie-cards and detected 26 p.Phe508del carriers, and a CF carrier incidence of 2.44%±0.75% (95% CI) (sample size calculator). About 1 in 37 people are CF carriers and the CF incidence is 1 in 5500. Conclusion: The frequency of the p.Phe508del in Portuguese CF patients 66.4% is similar to the French 67.2% and other European countries, but different from the reported data of Spain 50.6%. The CF incidence in our population of 1 in 5500 is much lower than expected, probably due to our population admixture.

11 A new multiplex PCR method for the quantification of aberrant transcripts from nasal epithelial cells of patients

Objective: The c.3909C>G (N1303K) is the most frequent mutation after the c.1521_1523delCTT (F508del) in Lebanon. A CFTR gene screening was performed on 10 Lebanese and 5 French CF patients carrying c.3909C>G. Parental studies revealed that all individuals have an association of c.3909C>G (exon 24) with c.869+11C>T (intron 7) in cis, inducing a new complex allele (CA). Since both mutations are located in two distant regions, we studied their combined impact by two plasmid constructions. Method: Firstly, we studied the impact of c.3909C>G on alternative splicing (AS), localization and maturation of the CFTR protein. Secondly, to study the impact of c.869+11C>T on AS, an ex-vivo study will be conducted using a plasmid containing the exon 7 and its flanking introns. After transfection of different type of eukaryotic cells, RNA extraction and RT-PCR, cDNA will be sequenced to verify the possible AS. Results: We showed that c.3909C>G affects the localization and the process. The in-silico study is in favor for a possible role of c.869+11C>T in AS. In fact, the used algorithm human splicing Finder showed that c.869+11C>T might affect the AS since it is located near the donor site. Conclusion: The class II mutation, c.3909C>G, induces mild phenotype with few CFTR on membrane. We suggest that the CA c.[869+11C>T;3909C>G], will act like a class I mutation, explaining the severe phenotype in some c.3909C>G patients. So, the classification of a mutation is not sufficient in a clinical approach, as the possible presence of a CA may alter its the specific effect. Therefore, an individualized approach is required to perform proper diagnostic and treatment when available.

5* Assessing the impact of unclassified variants on splicing of CFTR mRNA: in silico predictions versus ex vivo assays

5* Assessing the impact of unclassified variants on splicing of CFTR mRNA: in silico predictions versus ex vivo assays C. Raynal1,2,3, M. Taulan2,3, V. Viart2,3, C. Theze2,3, J. Miro2,3, S. TufferyGiraud2,3, M. Claustres1,2,3, M. des Georges1,3. 1CHU Montpellier, Laboratoire de Genetique des Maladies Rares, Montpellier, France; 2Universite Montpellier 1, Montpellier, France; 3INSERM − U827, Montpellier, France

Search for population-specific CFTR mutations: How to avoid the identification of unclassified variants

More than 1600 different sequence variations have been reported in the CFTR gene and the mutational spectrum varies in accordance with geographic and/or ethnic origins of patients. According to the recommendations from best practice guidelines, it is advisable to offer to patients/relatives/partners a mutation detection rate higher than 80% including population-specific mutations with relative frequencies above 1%. Additional mutation screening using scanning methods (DGGE, DHPLC, HRM) or sequencing should be performed to complete the search of the most frequent mutations by commercial kits, however they permitted the identification of rare/unclassified variants. As the results of a genetic test can have serious implications for an individual and his family, it is important to identify, particularly in carrier screening or in case of suspicion of CF in fetuses with bowel hyperechogenic, only CF-causing mutations. We designed 3 population-specific (North African, Spanish and Italian) home-made kits based on SNaPshot method (single nucleotide primer extension with dideoxy nucleotide, migration on 3130 Genetic Analyser and interpretation of results using GeneMapper Software from Applied Biosystem). According to literature data, each panel includes mutation subsets (6 to 9 different disease-causing mutations) shown to be frequent in French subjects with foreign origins and allows the identification of both wild type and mutant alleles. In order to avoid misinterpretation of unclassified missenses or putative splice mutations, only variants of proven or certain clinical relevance are included in these panels. This accurate, specific, sensible and rapid method can easily be adapted and developed to other population-specific mutations. Supported by: VLM