Targeted identification of reported deep-intronic variants in ABCA4 in 224 French Stargardt disease cases

Abstract

Purpose : In the last two decades, ABCA4 has been tested in 1087 French Stargardt disease (STGD1) cases by employing high-throughput analyses. Many remained unsolved because the two causal mutations in ABCA4 gene were not found. Recently, a frequent coding (p.Asn1868Ile) and a non-canonical splice site variant (c.5461-10T>C), were deemed causative. In addition, several causal deep-intronic variants were recently identified. Therefore, we aimed to search for these variants in 224 French STGD1 cases to improve the genetic diagnosis for patients in Lille.

Methods : We ascertained 135 monoallelic STGD1 probands and 89 STGD1 cases with no ABCA4 variants. The 50 ABCA4 exons and 12 selected intronic regions were sequenced using 513 single molecule inversion probes (smMIPs). The smMIPs were pooled, phosphorylated and hybridized to 100 ng of genomic DNA of each patient. Sequencing analysis was performed using one NextSeq 500 high output module. Novel non-canonical splice site and deep-intronic variants were tested for splice defects using in vitrosplice assays in HEK393T cells employing ABCA4 midigenes.

Results : smMIPs-based sequencing solved 71 (31.7%) STGD1 cases and found one ABCA4 variant in 63 (28.1%) probands. 31 (13.8%) probands carried p.(Asn1868Ile) in trans with a severe mutation and are likely solved. The latter showed a late age at onset (average 47.8 yrs) and consisted of more females (22/9, 70.9%) than males. In addition, we found 42 deep intronic variants in 40 (17.8%) probands of which c.4253+43G>A (n=17) and c.5196+1137G>A (n=13) were the most frequent. A novel deep-intronic variant, c.4539+2065C>G, showed a 170-nt pseudoexon insertion in the mRNA when tested in ABCA4 midigenes. Similarly, three novel non-canonical splice-site variants, c.4128G>C, c.4539G>A and c.4849G>A led to splice defects, i.e. a 12-nt exon 27 elongation and skipping of exon 30 and 35, respectively.

Conclusions : By employing a smMIPs-based sequencing method, we could (likely) solve ~45.5% of previously genotyped STGD1 probands. The high proportion (17.8%) of STGD1 cases with deep intronic variants highlights the relevance to investigate ABCA4i ntronic regions in standard genotyping. The smMIPs platform is very cost-effective and proved its efficiency for the analysis of coding and intronic regions. In follow-up studies, the entire ABCA4 gene will be sequenced employing smMIPs.

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.