A cryptic BRCA2 repeated motif binds to HSF2BP oligomers with no impact on meiotic recombination

BRCA2 plays a prominent role in meiotic homologous recombination (HR). Loss of BRCA2 or several of its meiotic partners causes fertility defects. One of these partners, HSF2BP, was recently discovered as expressed physiologically in germline and ectopically produced in cancer cells. It has an N-terminal coiled coil motif involved in direct binding to the protein BRME1, and both HSF2BP and BRME1 are essential for meiotic HR during spermatogenesis. It also interacts through its C-terminal Armadillo (ARM) domain with a conserved region of BRCA2 of unknown function. We analyzed the structural properties and functional consequences of the BRCA2-HSF2BP interaction and tested the emerging model of its involvement in meiosis. We solved the crystal structure of the complex between the BRCA2 fragment that is disordered in solution and the HSF2BP dimeric ARM domain. This revealed two previously unrecognized BRCA2 repeats that each interact with one ARM monomer from two different dimers. BRCA2 binding triggers ARM tetramerization, resulting in a complex containing two BRCA2 fragments connecting two ARM dimers. The 3D structures of the BRCA2 repeats are superimposable, revealing conserved contacts between the BRCA2 residues defining the repeats and the HSF2BP residues lining the groove of the ARM. This large interface is responsible for the nanomolar affinity of the interaction, significantly stronger than any other measured interaction involving BRCA2. Deleting exon 12 from Brca2 , encoding the first repeat, disrupted BRCA2 binding to HSF2BP in vitro and in cells. However, Brca2Δ12/Δ12 mice with the same deletion were fertile and did not show any meiotic defects, contrary to the prediction from the model positing that HSF2BP acts as a meiotic localizer of BRCA2. We conclude that the high-affinity interaction between BRCA2 and HSF2BP and the resulting HSF2BP oligomerization are not required for RAD51 and DMC1 recombinase localization to meiotic double strand breaks and for productive meiotic HR. ### Competing Interest Statement The authors have declared no competing interest.