Mild TSC phenotype and non-penetrance associated with a frameshift variant in TSC2 prompts caution in evaluating pathogenicity of frameshift variants

Technological advances in genetic testing, particularly the adoption of noninvasive prenatal screening (NIPS) for single gene disorders such as tuberous sclerosis complex (TSC, OMIM# 613254), mean that putative/possible pathogenetic DNA variants can be identified prior to the appearance of a disease phenotype. Without a phenotype, accurate prediction of variant pathogenicity is crucial. Here, we report a TSC2 frameshift variant, NM_000548.5(TSC2):c.4255_4256delCA, predicted to result in nonsense-mediated mRNA decay (NMD) and cessation of TSC2 protein production and thus pathogenic according to ACMG criteria, identified by NIPS and subsequently detected in family members with few or no symptoms of TSC. Due to the lack of TSC-associated features in the family, we hypothesized that the deletion created a non-canonical 5′ donor site resulting in cryptic splicing and a transcript encoding active TSC2 protein. Verifying the predicted effect of the variant was key to designating pathogenicity in this case and should be considered for other frameshift variants in other genetic disorders. Phenotypic information on the family members was collected via review of the medical records and patient reports. RNA studies were performed using proband mRNA isolated from blood lymphocytes for RT-PCR and Sanger sequencing. Functional studies were performed by transient expression of the TSC2 variant proteins in cultured cells, followed by immunoblotting. No family members harboring the variant met any major clinical diagnostic criteria for TSC, though a few minor features non-specific to TSC were present. RNA studies supported the hypothesis that the variant caused cryptic splicing, resulting in an mRNA transcript with an in-frame deletion of 93 base pairs r.[4255_4256del, 4251_4343del], p.[(Gln1419Valfs*104), (Gln1419_Ser1449del)]. Expression studies demonstrated that the canonical function of the resulting truncated TSC2 p.Gln1419_Ser1449del protein product was maintained and similar to wildtype. Although most frameshift variants are likely to result in NMD, the NM_000548.5(TSC2):c.4255_4256delCA variant creates a cryptic 5′ splice donor site, resulting in an in-frame deletion that retains TSC2 function, explaining why carriers of the variant do not have typical features of TSC. The information is important for this family and others with the same variant. Equally important is the lesson that predictions can be inaccurate, and that caution should be used when designating frameshift variants as pathogenic, especially when phenotypic information to corroborate testing results is unavailable. Our work demonstrates that functional RNA- and protein-based confirmation of the effects of DNA variants improves molecular genetic diagnostics.