Partner Choice in Homologous Recombination: Sister Chromatid Repair, LOH, and Translocations

Partner Choice in Homologous Recombination: Sister Chromatid Repair, LOH, and Translocations


By definition, homologous repair of damaged DNA requires a homologous sequence that can serve as the template for repair. We recently defined the “partner choice” for homologous repair in mammalian cells using molecular genetic approaches. We found that sister chromatids serve as repair templates at high frequency (Johnson et al. Nature. 1999), but that homologous sequences on both homologous and nonhomologous chromosomes are also used, albeit at a significantly reduced frequency. (For homolog recombination, see Moynahan and Jasin. Proc Natl Acad Sci USA. 1997; for heterolog recombination, see Richardson et al. Genes Dev. 1998.) The reduced frequency may be due to the lower probability of random collision between homologous sequences located on 2 different chromosomes within the nucleus. Sister chromatids, on the other hand, are in close proximity to each other.

  • the enzyme is conserved among fungi
  • the structure and mechanism of the fungal RNA triphosphatases are different from the human enzyme
  • the triphosphatase activity is essential for yeast cell growth
  • metazoan species encode no obvious homologues of the fungal enzyme
  • Sister chromatids — Precise Repair
  • Homologs — Potential for LOH
  • Heterologs — Potential for Translocations

Since sister chromatids are identical to each other, recombination with a sister chromatid is expected, in most cases, to be a precise way for repairing DNA damage. Recombination between 2 different chromosomes has the potential to alter the genetic information in a cell. For example, when homologs recombine, loss of heterozygosity (LOH) can result, in which information from one parental chromosome is replaced by information from the other parental chromosome. LOH is detrimental when deleterious mutations are uncovered — for example, those in tumor suppressor genes.

We are currently determining the extent of LOH that occurs when homologs recombine and how frequently homolog recombination is associated with crossing-over. When nonhomologous chromosomes recombine, there is the potential for even more dramatic alterations in genome sequence and structure than LOH. Strikingly, however, we found that in most recombination events between nonhomologous chromosomes, only a small amount of sequence information was transferred from the unbroken chromosome onto the broken chromosome. The remaining recombinants transferred a larger amount of sequence information, but in no instances were translocations or other genome rearrangements observed.

Project Members