DNA repair enzymes have evolved to protect our genomes from an onslaught of exogenous DNA damaging agents, endogenous free radicals and replication errors. Many of these enzymes exist as complex machines that couple multiple catalytic steps in the DNA repair process. We have determined the overall architecture of two such “machines” using a combination of X-ray crystallography and electron microscopy.

First, we show that the 340kDa Bloom’s complex consists of 4 proteins that couple the DNA unwinding activity of the BLM helicase with the decatenation activity of topoisomerase IIIα. EM and biochemical studies of a recombinant complex reveal how it can separate covalently intertwined DNA strands at the conclusion of homologous recombination. Mutations in BLM are associated with familial breast cancer families and the congenital disorder Bloom’s syndrome. Some of these mutations lie at the interface of BLM with other proteins in the catalytic complex, affecting its ability to function as a multi-subunit holoenzyme.

Second, we have used EM to determine the architecture of a 260kDa heterodimer important for the recognition of damaged replication forks. FANCM:FAAP24 contains two DNA binding regions that are located in proximity within the large complex despite being separated by over 1500 amino acids in the FANCM protein. FANCM is also mutated in cancers, and the leukemia prone disorder Fanconi Anaemia. We provide a model for how FANCM detects DNA damage and activates a ubiquitination signaling cascade that leads to repair of the damage.

There are several benefits and pitfalls in using EM to characterise large protein complexes. The large multi-subunit nature of complex DNA repair machines make them ideal targets for this emerging technology.