DNA repair mechanisms appear to be present in all organisms, and play a significant role in both cell survival and cancer risk. Recombinational repair has been shown to be an important repair mechanism in bacteria and in simple eucaryotes such as yeast. Recently this repair mechanism has also been shown to be important in mammalian cells (1). The primary goal of our research has been to identify and characterize genes and their encoded proteins that are important for recombinational repair. Initially we concentrated on genes from the yeast S. cerevisiae (2), and in the last few years we have focused on related human genes. In yeast, the RAD50 to RAD57 genes are involved both in the recombinational repair of DNA damage, including double-strand breaks, and in meiotic recombination. The Rad51 protein is the major DNA strand transfer protein, while the Rad55 and Rad57 proteins form a dimer that interacts with Rad51 and facilitates its recombinational activity. The Rad55 and Rad57 proteins share some sequence homology with each other and with Rad51.

In human cells there are at least six proteins that share limited sequence homology with the yeast and human Rad51 proteins. Five of these proteins (XRCC2, XRCC3, Rad51B, Rad51C and Rad51D) probably play an important role in recombinational repair. Our group originally identified and continues to characterize the RAD51C gene (also called RAD51L2) (3). We are also concentrating on characterizing protein-protein interactions within this group of Rad51-related proteins. The XRCC3 protein has been shown to interact with Rad51 in the yeast two-hybrid system and in co-immunoprecipitation experiments (4). Each member of this family has been shown to interact with one or more other members using two-hybrid experiments. We have also used a yeast three-hybrid system (a derivative of the two-hybrid system in which a third protein is expressed in non-fused form) to show that many of these interactions can occur simultaneously. Our work suggests that the proteins we are studying may be involved in a large complex or "recombinosome". We are in the process of attempting to confirm these interactions with experiments in mammalian cells, and also plan to determine the biological significance of these interactions. In addition, we have used the yeast two-hybrid system to identify human proteins outside the Rad51 family that appear to interact with Rad51C. We are currently characterizing a few of these candidate genes to confirm the interaction and to see how they may impact recombinational repair and carcinogenesis.