A moving target for drug discovery: Structure activity relationship and many genome (de)stabilizing functions of the RAD52 protein

Abstract

BRCA-ness phenotype, a signature of many breast and ovarian cancers, manifests as deficiency in homologous recombination, and as defects in protection and repair of damaged DNA replication forks. A dependence of such cancers on DNA repair factors less important for survival of BRCA-proficient cells, offers opportunities for development of novel chemotherapeutic interventions. The first drugs targeting BRCA-deficient cancers, poly-ADP-ribose polymerase (PARP) inhibitors have been approved for the treatment of advanced, chemotherapy resistant cancers in patients with BRCA1/2 germline mutations. Nine additional proteins that can be targeted to selectively kill BRCA-deficient cancer cells have been identified. Among them, a DNA repair protein RAD52 is an especially attractive target due to general tolerance of the RAD52 loss of function, and protective role of an inactivating mutation. Yet, the effective pharmacological inhibitors of RAD52 have not been forthcoming. In this review, we discuss advances in the state of our knowledge of the RAD52 structure, activities and cellular functions, with a specific focus on the features that make RAD52 an attractive, but difficult drug target.

Keywords: BRCA1; BRCA2; DNA double strand break repair; Homologous recombination; Pharmacological inhibition of protein-DNA interactions; RAD52; Synthetic lethality in cancer.

Figure 1.. Synthetic lethality in BRCA-deficient cancers.

A. Biallelic deficiency in tumor suppressors BRCA1 and BRCA2 is embryonically lethal. Heterozygous inactivating mutations, on the other hand, predispose carriers to a range of cancers via an increased chance of loss of heterozygosity (LOH). The resulting BRCA-deficient tumors are sensitive to ionizing radiation and DNA damaging chemotherapeutics, but often develop resistance. B. BRCA-deficiency makes the cancer cell vulnerable to depletion and/or inhibition of DNA repair proteins with whom it has synthetically lethal relationships. The concept of synthetic lethality has been first applied to PARP1, but is now extended to nine additional proteins including RAD52. These nine proteins represent attractive anticancer drug targets and are depicted alongside the references to papers that identified each synthetic interaction, found a pharmacological inhibitor, or a DNA repair deficiency that exacerbates or rescues the synthetic lethal interaction (see text for details).

Figure 2.. RAD52 structure and interactions.

A. The RAD52 domain organization includes the well-structured N-terminal self-oligomerization domain (residues 1–209), which interacts with ssDNA and dsDNA, and largely disordered C-terminal domain involved in protein-protein interactions including RPA (orange) and RAD51 (blue), and also contains nucleo-localization sequence (NLS). The disorder probability along the RAD52 sequence was calculated using prDOS [132]. B. and C. Residues corresponding to the yeast separation of function mutants [75] are mapped on the PDB: 5XRZ structure [72]. D. Residues comprising primary (teal) and secondary (deep purple) DNA binding sites. E. Two lysine residues in the secondary DNA binding site flank tyrosine 104, which is phosphorylated by c-ABL kinase during DNA damage response, resulting in activation of the RAD52 annealing activity.

Figure 3.. Genome (de)stabilizing mechanisms that utilize RAD52.

Six distinct cellular functions have been attributed to RAD52 so far. While originally, RAD52 was thought to function mostly during S-phase, its newly identified functions take place in many phases of the cell cycle. The known RAD52 functions include: (A.) gatekeeper of stalled and damaged DNA replication forks, (B.) fork cleavage, (C.) annealing of complementary DNA strands in the presence of RPA leading to mutagenic SSA, (D.) antagonizing POLθ-mediated end joining, (E.) initiation of MiDAS, and (F.) inverse strand exchange leading to RNA-dependent DNA repair in G1/G0 phase.