Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

DNA repair

Making the cut

Analysis of the first step in repairing double-stranded-DNA breaks reveals that the Mre11 enzyme makes a DNA nick at a point separate from the break ends, creating an entry site for further processing by exonuclease enzymes. See Letter p.122

Double-stranded breaks in chromosomes are dangerous lesions that form when both strands of the DNA duplex are severed. Such breaks must be accurately repaired to preserve genome integrity — inaccurate or failed repair can result in chromosome rearrangements, loss of genetic information or even cell death. Indeed, faulty repair of double-stranded breaks is associated with infertility, developmental and immunological defects and predisposition to cancer. Cells repair breaks in an error-free manner through a mechanism called homologous recombination, which begins with removal of one of the two DNA strands at each broken end. On page 122 of this issue, Cannavo and Cejka1 provide mechanistic insight into exactly how a three-protein complex, Mre11–Rad50–Xrs2, is involved in the initial stages of error-free DNA repair.

The first step in homologous recombination is the degradation of the 5′ end of DNA on either side of a break to yield 3′ single-stranded DNA (ssDNA) tails — a process called end resection. The Rad51 protein then binds to these tails and promotes exchange of genetic information with homologous sequences from the sister chromosome, leading to DNA repair. Genetic studies2,3 in the budding yeast Saccharomyces cerevisiae suggest a two-step mechanism for end resection. Initially, the evolutionarily conserved Mre11–Rad50–Xrs2 (MRX) enzymatic complex and the Sae2 protein clip off the 5′-terminated DNA strand, creating a short 3′ overhang. Next, the overhang is rapidly lengthened by the Exo1 or Dna2 nuclease enzymes. These enzymes remove nucleotides from the strand being processed to generate an extensive tract of ssDNA.

Significant progress has been made in deciphering the mechanisms used by Exo1 and Dna2 for extensive resection4,5,6, but little is known about how MRX and Sae2 cooperate to initiate the process. Mre11 has exonuclease activity (it degrades DNA from the end of the strand), but an in vitro study7 showed that the enzyme catalyses DNA degradation from 3′ to 5′, the opposite direction to that in which end resection occurs. Sae2 also reportedly shows nuclease activity in vitro8. A bidirectional model9 proposes that MRX uses its 3′–5′ exonuclease activity to proceed back to the double-stranded break from an internal nick in the DNA created by MRX and Sae2 (Fig. 1). However, the identity of the endonuclease that could create this internal nick has been unclear.

Figure 1: Resection in the right direction.

Cannavo and Cejka1 examined the role of the Mre11–Rad50–Xrs2 (MRX) enzyme complex and Sae2 protein in end resection, the first step in repairing double-stranded breaks in DNA. Using a protein block to mimic a natural DNA break, they found that MRX makes a nick in the DNA at 15 to 20 nucleotides (nt) from the 5′ end of the break, a process that is promoted by phosphorylated (P) Sae2. The nick creates an entry site for Mre11 to proceed back to the break end in the 3′ to 5′ direction, and for the Exo1 or Dna2 enzymes to cleave 5′ to 3′, extending the resected end to leave a 3′ single-stranded DNA tail.

Cannavo and Cejka purified MRX and Sae2 and analysed the proteins' activities on model DNA substrates in vitro. They observed that MRX did indeed have 3′–5′ exonuclease activity. But, in contrast to previous work8, they found no nuclease activity for Sae2 alone.

The authors incubated MRX and Sae2 with a linear double-stranded DNA substrate in which one end was blocked by protein to mimic a double-stranded break. They observed an Sae2-dependent degradation product indicative of a nick internal to the protein-blocked DNA end. The researchers showed that this previously undocumented endonuclease activity was inherent to Mre11 — they repeated the experiment using a variant of the MRX complex containing a nuclease-defective version of Mre11, and observed no endonuclease activity.

Binding and hydrolysis of ATP molecules cause large conformational changes in the Rad50 subunit of the MRX complex, activating Mre11 nuclease activity10,11. Cannavo and Cejka observed MRX- and Sae2-dependent clipping activity only in the presence of ATP. This was eliminated using MRX mutants that were unable to bind or hydrolyse ATP. By adding protein blocks at both ends, the authors prevented the 3′–5′ exonuclease action of Mre11 and, using labelling techniques, they then mapped the site at which the enzyme clipped the DNA to around 15 to 20 nucleotides from the break ends. Thus, the properties of the in vitro reaction reported by Cannavo and Cejka match the known requirements for homologous recombination in vivo2,3. Furthermore, this study explains how Mre11 promotes resection of the 5′-terminated strand.

One key question is how Sae2 regulates MRX endonuclease activity. Phosphate molecules, which can modify protein behaviour, are added to Sae2 by a protein-kinase enzyme through phosphorylation when cells enter the S phase of the cell cycle. Mutation of a serine amino-acid residue (serine 267) to an alanine residue that cannot be phosphorylated severely impairs Sae2 function in vivo12. Cannavo and Cejka found that this mutant protein rendered MRX unable to clip double-stranded DNA. In addition, they showed that dephosphorylation of Sae2 with a phosphatase enzyme decreased its activity. These data suggest that Sae2 acts as a phosphorylation-dependent switch to trigger MRX endonuclease activity.

Because MRX clipping is dependent on ATP, and a specific point mutation in the RAD50 gene, like loss of Sae2, prevents DNA clipping13, it is likely that Sae2 acts through Rad50 to activate Mre11. Cannavo and Cejka detected a physical interaction between Sae2 and MRX, but found that, instead of the anticipated interaction with Rad50, only the Mre11 and Xrs2 subunits interacted with Sae2.

The CtIP protein is considered to be the equivalent of Sae2 in vertebrate cells and plays a crucial part in end resection. Two papers published earlier this year reported that CtIP has nuclease activity14,15. However, the protein's active site is not in its evolutionarily conserved carboxy-terminal domain, which, as the current study shows, is required in Sae2 to stimulate the latent MRX endonuclease. This raises the question of whether CtIP has further functions and can process DNA independently of the vertebrate form of MRX.

Cannavo and Cejka's study provides mechanistic insight into how MRX and Sae2 initiate end processing, but leads to more questions. For example, how does MRX recognize a protein-blocked end? How does a blocked end trigger cleavage of the 5′ strand? Finally, there is the question of how Sae2 phosphorylation coordinates with the ATPase activity of MRX to activate Mre11 endonuclease activity.


  1. 1

    Cannavo, E. & Cejka, P. Nature 514, 122–125 (2014).

    ADS  CAS  Article  Google Scholar 

  2. 2

    Mimitou, E. P. & Symington, L. S. Nature 455, 770–774 (2008).

    ADS  CAS  Article  Google Scholar 

  3. 3

    Zhu, Z., Chung, W.-H., Shim, E. Y., Lee, S. E. & Ira, G. Cell 134, 981–994 (2008).

    CAS  Article  Google Scholar 

  4. 4

    Cannavo, E., Cejka, P. & Kowalczykowski, S. C. Proc. Natl Acad. Sci. USA 110, E1661–E1668 (2013).

    ADS  CAS  Article  Google Scholar 

  5. 5

    Niu, H. et al. Nature 467, 108–111 (2010).

    ADS  CAS  Article  Google Scholar 

  6. 6

    Cejka, P. et al. Nature 467, 112–116 (2010).

    ADS  CAS  Article  Google Scholar 

  7. 7

    Paull, T. T. & Gellert, M. Mol. Cell 1, 969–979 (1998).

    CAS  Article  Google Scholar 

  8. 8

    Lengsfeld, B. M., Rattray, A. J., Bhaskara, V., Ghirlando, R. & Paull, T. T. Mol. Cell 28, 638–651 (2007).

    CAS  Article  Google Scholar 

  9. 9

    Garcia, V., Phelps, S. E. L., Gray, S. & Neale, M. J. Nature 479, 241–244 (2011).

    ADS  CAS  Article  Google Scholar 

  10. 10

    Lammens, K. et al. Cell 145, 54–66 (2011).

    CAS  Article  Google Scholar 

  11. 11

    Lim, H. S., Kim, J. S., Park, Y. B., Gwon, G. H. & Cho, Y. Genes Dev. 25, 1091–1104 (2011).

    CAS  Article  Google Scholar 

  12. 12

    Huertas, P., Cortés-Ledesma, F., Sartori, A. A., Aguilera, A. & Jackson, S. P. Nature 455, 689–692 (2008).

    ADS  CAS  Article  Google Scholar 

  13. 13

    Alani, E., Padmore, R. & Kleckner, N. Cell 61, 419–436 (1990).

    CAS  Article  Google Scholar 

  14. 14

    Wang, H. et al. Mol. Cell 54, 1012–1021 (2014).

    CAS  Article  Google Scholar 

  15. 15

    Makharashvili, N. et al. Mol. Cell 54, 1022–1033 (2014).

    CAS  Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Lorraine S. Symington.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Symington, L. Making the cut. Nature 514, 39–40 (2014).

Download citation

Further reading


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing