Integrated RNA-based functional characterization facilitates reclassification of BRCA1/2 variants of uncertain significance in a clinical cancer cohort

Background: Variants of uncertain significance in BRCA1 and BRCA2 remain a major challenge in hereditary breast and ovarian cancer genetic testing, particularly when their effects on pre-mRNA splicing are unclear. Computational prediction can facilitate the identification of potentially splice-altering variants but may yield discordant results and cannot fully resolve their clinical significance. This study aimed to characterize the splicing consequences of selected BRCA1/2 variants and evaluate the contribution of functional RNA evidence to variant interpretation. Results: Sixty unique BRCA1/2 variants of uncertain significance identified among 1,762 patients with cancer were evaluated for splice-altering potential, and five candidate variants were selected for functional analysis. Minigene assays identified two splice-altering BRCA2 variants, c.476-3C > A and c.8332-13T > G, whereas BRCA2 c.632 − 18_632-17del and c.7435 + 20G > A and BRCA1 c.4676-6T > C predominantly retained normal splicing. The c.476-3C > A variant predominantly induced exon skipping. In contrast, c.8332-13T > G generated a more complex splice pattern comprising exon skipping and activation of two cryptic 3’ splice acceptor sites. Exon skipping caused by c.8332-13T > G resulted in the in-frame deletion p.(Ile2778_Gln2829del). RNA sequencing further characterized these aberrant splice-junction patterns and confirmed the complex transcript profiles observed in the minigene assays. Structural analysis and molecular dynamics simulations were subsequently used to characterize the native BRCA2 – DSS1 structural environment surrounding the region deleted in p.(Ile2778_Gln2829del). Familial segregation provided additional supportive evidence for the c.8332-13T > G variant. Integration of functional RNA findings with predicted protein consequences and familial evidence supported classification of c.476-3C > A and c.8332-13T > G as likely pathogenic and pathogenic, respectively, while the three splice-neutral variants were classified as likely benign. Conclusions: Experimental RNA analysis can resolve BRCA1/2 variants that remain difficult to interpret from sequence information alone. In this study, combined minigene analysis, transcript quantification, and RNA sequencing revealed complex variant-specific splice patterns and provided functional evidence supporting reclassification of two BRCA2 variants as pathogenic and three splice-neutral variants as likely benign. These findings also demonstrate that disruption of normal splice-site recognition can redistribute usage among competing splice sites and generate multiple transcripts with distinct molecular consequences. Integrating RNA-based functional evidence with predicted protein consequences and familial data may therefore improve the clinical interpretation of selected BRCA1/2 variants of uncertain significance.