When a DNA molecule breaks, its complementary copy can be used as a template for repair. A familiar protein complex is recruited to the damaged site, keeping it close to the undamaged copy.
Before a cell divides, it must replicate its DNA, producing two identical copies of each chromosome, known as sister chromatids. When one of the sister chromatids suffers a break that affects both of its DNA strands, it is mended by ‘homologous recombination’. In this process, information on the undamaged sister is used as a template for repair. This mechanism is essential for cell survival and genome stability. But how can the repair machinery find the proper template in the crowded environment of the cell nucleus? Writing in Molecular Cell, Ünal et al.1 and Ström et al.2 provide compelling evidence that a protein complex called cohesin has a crucial role.
The cohesin complex was originally identified as a protein component that ensures the proper segregation of sister chromatids during cell division. It does so by holding them together from the time that they are produced (during the ‘S phase’ of the cell-division cycle) until the mid-stage of mitosis — the phase in which the sister chromatids of each pair are separated in preparation for making two cells3.
However, previous genetic studies also implicated cohesin in the repair of double-strand breaks (DSBs) between S phase and mitosis (in the post-replicative, or G2, phase). In fact, one of the cohesin sub-units had been found to be involved in DSB repair in fission yeast4 long before its essential role in chromosome segregation was recognized.
Although subsequent studies in budding yeast5 and vertebrates6 also supported a requirement for cohesin in post-replicative DSB repair, it remained unclear how the complex participates in this process at a mechanistic level. In principle, two models can be considered. During S phase, cohesin complexes — spaced 10–15 kilobases apart in yeast — build up a physical linkage along the length of the sister chromatids; this genome-wide linkage may be sufficient to keep sisters close enough to allow repair by homologous recombination when a DSB appears in G2 phase. Alternatively, cohesin may have a more active, more specialized and more local role around the damaged sites.
To follow the dynamics of cohesin in response to a DSB in the budding yeast Saccharomyces cerevisiae, Ünal et al.1 and Ström et al.2 have used a well-characterized system in which a DSB can be generated at a defined point of the genome, by inducing the expression of a specific DNA-cutting enzyme called HO endonuclease. By using a high-resolution mapping technique — chromatin immunoprecipitation — both groups find that cohesin accumulates in a large chromosomal region of around 50–100 kilobases surrounding the DSB. Moreover, this local enrichment of cohesin at the DSB site occurs in G2 phase and requires the same protein (Scc2) that facilitates genome-wide loading of cohesin in early S phase. These observations strongly suggest that the DSB induces de novo loading of cohesin at the damaged site, rather than a rearrangement of cohesin already loaded on the DNA during S phase.
What is the molecular mechanism underlying this local recruitment of cohesin? A previous study7 showed that H2AX — a histone protein, involved in packaging DNA — is phosphorylated in a large region that extends some 50 kilobases either side of an HO-endonuclease-induced DSB in yeast. This modification depends on two enzymes — namely Tel1 and Mec1 (ATM and ATR, respectively, in mammals) — which are involved in a ‘checkpoint’ that delays cell division until damaged DNA is repaired7.
Following on from these findings, Ünal et al.1 show that cohesin recruitment in response to a DSB is also regulated by Tel1 and Mec1, and requires H2AX phosphorylation. The formation of the region of phosphorylated H2AX proteins is therefore likely to be a prerequisite for the loading of cohesin (Fig. 1). Equally important, Ström et al.2 provide evidence that the DSB-induced cohesin loading does indeed establish a de novo linkage between the damaged chromatid and its undamaged sister, thereby facilitating DSB repair. Consistent with this conclusion, cohesin is not required for other repair-related processes, such as intrachromosomal gene conversion or the resection of broken DNA strands1 — events that do not require a template or, therefore, new sister-chromatid linkages.
These observations made in yeast are likely to be highly relevant to our understanding of DSB repair in mammals. In fact, DNA-damage-induced recruitment of cohesin was first hinted at through a lower-resolution, cytological method — immunofluorescent microscopy — in human cells8. This recruitment was dependent on the DSB-repair protein Mre11, as has been confirmed in yeast1. Moreover, other studies showed that ATM phosphorylates one component of cohesin following DNA damage induced by ionizing irradiation, and that this phosphorylation is required for activation of the DNA-damage checkpoint9,10. Cells expressing only a non-phosphorylatable form of cohesin support normal mitotic progression, but exhibit a defective checkpoint response and increased chromosomal aberrations11, implying that this modification of cohesin is specialized for DNA repair in mammalian cells.
In summary, the new studies1,2 underscore the importance of a cohesin-mediated sister linkage in DSB repair. They also provide a fresh view of cohesin — which seems to be much more dynamic than had been thought — and raise several new questions. What is the real signal or modification that allows cohesin to accumulate at the DSB site? Does cohesin, or the protein that loads it onto DNA, interact directly with phosphorylated H2AX? What happens when DNA repair is completed? Might cohesin need to be cleaved in order to be unloaded after DNA repair12, as has been shown in mitosis3? Undoubtedly, answering these questions will further enrich our understanding of the global and local dynamics of chromosome architecture, and its impact on genome stability.
Ünal, E. et al. Mol. Cell 16, 991–1002 (2004).
Ström, L., Lindroos, H. B., Shirahige, K. & Sjögren, C. Mol. Cell 16, 1003–1015 (2004).
Nasmyth, K. Annu. Rev. Genet. 35, 673–745 (2001).
Birkenbihl, R. P. & Subramani, S. Nucl. Acids Res. 20, 6605–6611 (1992).
Sjögren, C. & Nasmyth, K. Curr. Biol. 11, 991–995 (2001).
Sonoda, E. et al. Dev. Cell 1, 759–770 (2001).
Shroff, R. et al. Curr. Biol. 14, 1703–1711 (2004).
Kim, J. S., Krasieva, T. B., LaMorte, V., Taylor, A. M. & Yokomori, K. J. Biol. Chem. 277, 45149–45153 (2002).
Kim, S.-T., Xu, B. & Kastan, M. B. Genes Dev. 16, 560–570 (2002).
Yazdi, P. T. et al. Genes Dev. 16, 571–582 (2002).
Kitagawa, R., Bakkenist, C. J., McKinnon, P. J. & Kastan, M. B. Genes Dev. 18, 1423–1438 (2004).
Nagao, K., Adachi, Y. & Yanagida, M. Nature 430, 1044–1048 (2004).
About this article
PLoS Genetics (2013)
Fission Yeast Cut8 Is Required for the Repair of DNA Double-Strand Breaks, Ribosomal DNA Maintenance, and Cell Survival in the Absence of Rqh1 Helicase
Molecular and Cellular Biology (2007)
BRCA1 Regulates RAD51 Function in Response to DNA Damage and Suppresses Spontaneous Sister Chromatid Replication Slippage: Implications for Sister Chromatid Cohesion, Genome Stability, and Carcinogenesis
Cancer Research (2005)