Investigation of the BRCT repeats in human hereditary breast cancer and DNA damage response

Abstract

The C-terminal region of breast cancer susceptibility gene 1 (BRCA1) contains a pair of tandem BRCT repeats that are critical for the tumour suppressor function of BRCA1. BRCT repeats are present in a large of number of proteins that are implicated in the cellular response to DNA damage. A subset of tandem BRCT domains, including those of BRCA1, functions as phosphorecognition modules. Aside from BRCA1, the precise molecular mechanisms of the BRCT repeats of other proteins remain largely unknown. We determined the crystal structure of the tandem BRCT domain of human mediator of DNA checkpoint 1 (MDC1) at 1.45 Å resolution. Our structural and biochemical studies suggest that the tandem BRCT domain of MDC1 functions as the predominant histone variant, γH2AX phosphorecognition module and that the interaction is critically dependent on the free carboxylate group of the γH2AX C-terminal tail. We also determined the crystal structure of the tandem BRCT domain of human BARD1, the in vivo binding partner of BRCA1. Our structure uncovers a degenerate phosphopeptide binding pocket that lacks the key arginine critical for phosphopeptide interactions in other BRCT proteins. Our biochemical studies reveal that a flexible tether links ankyrin and BRCT domains in BARD1. Furthermore, the linker is required for the interactions between the CstF-50 WD-40 domain and BARD1, allowing the BARD1 C-terminus to convey DNA damage signals directly to RNA polymerase. Finally, using protease-based and phosphopeptide pull-down assays, we directly assessed the structural and functional effects of 117 single amino acid substitutions in the BRCA1 BRCT domain derived from breast cancer screening programs. None of the variants showing enhanced sensitivity to proteolytic digestion were found to be active in peptide binding, indicating that these missense mutations contribute to BRCA1 loss of function through protein destabilizing effects. A subset of structurally stable variants was defective in peptide binding activity, suggesting that these variants may disrupt the phosphopeptide binding pocket. Taken together, the results reveal that 32% of the variants show structural stability and peptide binding activity that were indistinguishable from those of wild type.