Refinement of the assignment to the ACMG/AMP BS3 and PS3 criteria of eight BRCA1 variants of uncertain significance by integrating available functional data with protein interaction assays

The clinical screening of cancer predisposition genes has led to the identification of a large number of variants of uncertain significance (VUS). Multifactorial likelihood models that predict the odds ratio for VUS in favor or against cancer causality, have been developed, but their use is limited by the amount of necessary data, which are difficult to obtain for rare variants. The guidelines for variant interpretation of the American College of Medical Genetics and Genomics along with the Association for Molecular Pathology (ACMG/AMP) state that "well-established" functional studies provide strong support of a pathogenic or benign impact (criteria PS3 and BS3, respectively) and can be used as evidence type to reach a final classification. Moreover, the Clinical Genome Resource Sequence Variant Interpretation Working Group developed rule specifications to refine the PS3/BS3 criteria. Recently, Lira PC et al. developed the "Hi Set" approach that generated PS3/BS3 codes for over two-thousands BRCA1 VUS. While highly successful, this approach did not discriminate a group of variants with conflicting evidences. Here, we aimed to implement the outcomes of the "Hi-set" approach applying Green Fluorescent Protein (GFP)-reassembly assays, assessing the effect of variants in the RING and BRCT domains of BRCA1 on the binding of these domains with the UbcH5a or ABRAXAS proteins, respectively. The analyses of 26 clinically classified variants, including 13 tested in our previous study, showed 100% sensitivity and specificity in identifying pathogenic and benign variants for both the RING/UbcH5a and the BRCTs/ABRAXAS interactions. We derived the strength of evidences generated by the GFP-reassembly assays corresponding to moderate for both PS3 and BS3 criteria assessment. The GFP-reassembly assays were applied to the functional characterization of 8 discordant variants from the study by Lyra et al. The outcomes of these analyses, combined with those reported in the "Hi Set" study, allowed the assignment of ACMG/AMP criteria in favor or against pathogenicity for all 8 examined variants. The above findings were validated with a semi-quantitative Mammalian Two-Hybrid approach, and totally concordant results were observed. Our data contributes in shedding light on the functional significance of BRCA1 VUS and on their clinical interpretation within the ACMG/AMP framework.