Plot of the HR efficiencies measured after expression of either WT BRCA1 or VUS normalized to the HR efficiency after expression of WT BRCA1. The HR efficiencies are here plotted as a function of the position of the mutation in the BRCA1 sequence. Statistical significance was calculated using a one-side paired student t-test. To account for the multiple testing, the p-values were adjusted using the Benjamini-Horchberg method at a level a=0.05. HR efficiencies significantly different from the WT value are marked by asterisks. They are indicated by * if p < 0.05 and ** if p < 0.01. All 7 (likely) causal (noted LC or C) variants (in red) caused a significantly decreased HR efficiency whereas all 6 (likely) neutral (noted LN or N) variants (in blue) caused no significant HR difference. Twenty-four VUS of class 3 could not provide WT HR efficiency.
ARTICLE ABSTRACT
BRCA1 mutations have been identified that increase the risk of developing hereditary breast and ovarian cancers. Genetic screening is now offered to patients with a family history of cancer, to adapt their treatment and the management of their relatives. However, a large number of BRCA1 variants of uncertain significance (VUS) are detected. To better understand the significance of these variants, a high-throughput structural and functional analysis was performed on a large set of BRCA1 VUS. Information on both cellular localization and homology-directed DNA repair (HR) capacity was obtained for 78 BRCT missense variants in the UMD-BRCA1 database and measurement of the structural stability and phosphopeptide-binding capacities was performed for 42 mutated BRCT domains. This extensive and systematic analysis revealed that most characterized causal variants affect BRCT-domain solubility in bacteria and all impair BRCA1 HR activity in cells. Furthermore, binding to a set of 5 different phosphopeptides was tested: all causal variants showed phosphopeptide-binding defects and no neutral variant showed such defects. A classification is presented on the basis of mutated BRCT domain solubility, phosphopeptide-binding properties, and VUS HR capacity. These data suggest that HR-defective variants, which present, in addition, BRCT domains either insoluble in bacteria or defective for phosphopeptide binding, lead to an increased cancer risk. Furthermore, the data suggest that variants with a WT HR activity and whose BRCT domains bind with a WT affinity to the 5 phosphopeptides are neutral. The case of variants with WT HR activity and defective phosphopeptide binding should be further characterized, as this last functional defect might be sufficient per se to lead to tumorigenesis.
The analysis of the current study on BRCA1 structural and functional defects on cancer risk and classification presented may improve clinical interpretation and therapeutic selection.
