It is impossible to avoid DNA damage. Sunlight, natural substances in food, and chemicals that are normally produced in our cells all can damage DNA. It is the DNA structure that provides the information necessary for life, so it is critical that damage be efficiently repaired in order to preserve genetic information.

The ability to keep DNA sequences intact must have been one of the very first accomplishments of early life forms as they battled against a harsh environment.  However, evolution is clearly a balance between maintaining DNA sequences and allowing for improvements via shuffling of DNA sequences, such as during meiosis.  Homologous recombination serves an important function during meiosis, since it allows for diversification of offspring

Although essential for meiosis, it is likely that homologous recombination actually did not evolve for the purpose of diversification, but rather for the purpose of stability.  Homologous recombination is a critical DNA repair pathway in normal somatic cells, functioning to help repair breaks and gaps that arise during DNA replication (Figure 1).  Thus, recombinational repair pathways are thought to have evolved long before meiosis.

Figure 1.  The three major classes of DNA damage that are repaired or tolerated using homologous recombination are A) a two-ended double strand break, B) a one-ended double strand break or broken replication fork, and C) a daughter strand gap.

Homologous recombination is critical during S phase, when it is used for repairing broken replication forks and daughter strand gaps.  Broken replication forks are one-ended double strand breaks that can be repaired by reinsertion of the broken end.  When breaks are comprised of a single end, they are considered to be 'one ended'.  Two ended double strand breaks can also be repaired by homologous recombination (although in mammals, two-ended breaks are more frequently repaired by non-homologous end-joining).  Repair of two-ended breaks is thought to occur either by the prototypic model (originally proposed by Szostak et al.), or by synthesis dependent strand annealing (SDSA) (Figure 2). 

Figure 2.  The prototypic model of repair of a two-ended double strand break is shown in A-E (as originally described by Szostak et al.).  Part F shows the synthesis dependent strand annealing (SDSA) model described by Haber.

Homology directed repair is essential for vertebrate cells.  Indeed, mammalian cells deficient in homologous recombination accumulate chromosome aberrations, which may be the result of mis-repair of one-ended double strand breaks via non-homologous end-joining.  Consistent with this observation is the finding that there people who inherit a deficiency in the ability to properly perform homologous recombination have an increased risk of cancer.  Thus, too little homologous recombination can promote cancer.

Although homologous recombination is essential, it is also possible that too much homologous recombination promotes cancer.  Although generally accurate, there is a finite risk of misalignment during homologous recombination, and this can promote tumorigenic sequence rearrangements (e.g., deletions and insertions) and contribute to loss of heterozygosity (caused by replacement of the sequence on one homologous chromosome with sequences from the other homologous chromosome).

Most genes known to modulate the frequency of homologous recombination in mammals are associated with cancer.  Thus, it is clear that genomic stability, homologous recombination, and cancer are closely linked.

Given the importance of this repair process, we set out to create the first mouse model in which cells that have undergone a homologous recombination event can be directly detected by emission of a fluorescent signal, and to use these mice to learn more about how environmental and genetic factors modulate homologous recombination in mammals in vivo.