Identification and molecular mechanisms of pathogenic splicing variants in neuromuscular disorders

Abstract

Background For families with rare Mendelian disorders, obtaining a precise genetic diagnosis is essential to enable personalised and preventative medicine. However, only ∼50% of families receive a genetic diagnosis following genetic testing, as there are still many challenges in the clinical interpretation of genetic variation. One such challenge is the identification and classification of variants that disrupt precursor messenger RNA (pre-mRNA) splicing, hereafter termed splicing variants. Even though splicing variants are well recognised as a common cause of Mendelian disorders, they are regularly classified as Variants of Uncertain Significance (VUS) rendering them clinically unactionable. It is our inability to accurately predict if and how a variant disrupts pre-mRNA splicing that prevents definitive classification of putative splicing variants.

Aims To explore the molecular mechanisms by which variants disrupt pre-mRNA splicing, thereby improving our ability to detect and accurately assign pathogenicity to splicing variants in the context of genetically diagnosing individuals with neuromuscular disorders.

Methods We have a cohort of 214 families with neuromuscular disorders for whom we have iteratively applied exomic, genomic and transcriptomic diagnostic sequencing approaches. Putative splicing variants were identified within this cohort or referred to us as VUSs from collaborators. Functional studies were performed to assess variant effect on pre-mRNA splicing. Analysis of datasets derived from variant databases and the human reference genome provided additional insights.

Results Numerous complex splicing variants were investigated and established as pathogenic for a variety of neuromuscular conditions. Splicing variants analysed included intronic deletions, deep intronic variants, coding variants, structural rearrangements, and extended splice site variants. The mechanistic basis of aberrant pre-mRNA splicing arising from these variants were explored, revealing novel or under-recognised disease mechanisms.

Conclusions We reveal important mechanistic insights behind pre-mRNA splicing that can be incorporated into diagnostic pipelines and used to inform new splicing prediction algorithms. The discoveries and lessons learned within this thesis are applicable to all rare Mendelian disorders and cancer genomics.