Novel Structural Variants Originating in F8 Non-Coding Regions in Previously Unresolved Cases of Severe Hemophilia A

Background. Hemophilia A is a rare X-linked bleeding disorder resulting from deficiency in coagulation factor VIII. Numerous genetic variants (>2000) affecting the F8 gene have been implicated as causative of hemophilia A, including structural variants (SVs) such as copy number variants (CNVs) and large intra-chromosomal inversions caused by recombination between distant regions with high homology to sequences within F8 intron 1 or intron 22. SVs detected in patients with hemophilia are associated with more severe disease, and different types of SVs may inform inhibitor risk. For the vast majority of patients, causative variants can be identified using targeted DNA sequencing of F8 coding regions and/or the use of methods which detect known SVs (e.g. inverse shifting PCR, long-range PCR, MLPA). However, these approaches fail to explain 1-3% of hemophilia A cases. We hypothesized that a dedicated structural variant analysis at the F8 locus using whole genome sequencing data could identify previously undetected deleterious F8 gene variants in unsolved cases of hemophilia A.

Methods. Cases were selected from the My Life, Our Future (MLOF) hemophilia study cohort recently whole genome sequenced by the NHLBI TOPMed program. In this study, we performed a custom SV analyses using whole genome sequencing (WGS) data from 11 cases of severe hemophilia A (factor VIII activity level < 1%) that remained genetically unexplained after exhausting all available laboratory testing methods. Two of the eleven unsolved severe hemophilia A cases (18%) were reported to have had an inhibitor.

Results. SV analyses of the F8 genomic region revealed previously undetected deletions and inversions in 6 out of the 11 cases. In these 6 samples, SV calls were supported by multiple sequencing reads (> 25 reads) and multiple types of read evidence (read depth, paired-end and/or split read evidence). Two deletions within intron 6 were detected in a single hemophilia A case, a finding which suggests F8 intron 6 may contain one or more regulatory elements critical for F8 expression. Three distinct large inversions predicted to disrupt the F8 structural gene were detected in five other cases; a 720Kb inversion with breakpoints in F8 intron 6 and SPRY3 intron 1 (n=1), a 20Mb inversion with breakpoints in F8 intron 1 and INTS6L intron 8 (n=1), and a 7.4Kb inversion with breakpoints in F8 intron 25 and the SMIM9 intron 1 (n=3). These events are novel in hemophilia and were also not present in the larger, sequenced My Life, Our Future dataset (N=2186), supporting these SVs as likely causative of severe hemophilia A. Both cases with inhibitors had the F8 intron 25-SMIM9 inversion.

Conclusions. This work demonstrates that dedicated analyses of WGS for SVs originating in non-coding regions can identify novel variants in previously unsolved cases of hemophilia A. We conclude that any genetic studies of diseases caused by loss-of-function variants should consider dedicated analyses for SVs. We predict additional deleterious SVs remain to be discovered in rare unexplained cases of hemophilia.