The protein Sae2 mediates the repair of double-strand breaks in DNA. It emerges that Sae2 activity is controlled by both its modification with acetyl groups and its degradation by the process of autophagy. See Article p.74
Cells use myriad ways to regulate the complex processes involved in their function. To control protein activity and stability, for example, an oft-used mechanism is post-translational modification of the protein. On page 74 of this issue, Robert et al.1 report one such modification that links the seemingly unrelated processes of DNA-damage repair and autophagy. Their observations simultaneously highlight the depth of cellular ingenuity and the immense interconnectedness of biological pathways.
The authors began by examining the effect of a specific post-translational modification — protein acetylation, in which an acetyl group is added to a protein. They used the drug valproic acid (VPA) to inhibit histone deacetylase (HDAC) enzymes, thereby causing hyperacetylation of histone proteins and reduced HDAC activity2. This treatment had no effect on cells, but after exposure to various DNA-damaging agents, the apparently normal VPA-treated cells were unable to activate the typical response to DNA damage.
Robert et al. present several lines of evidence to explain the failure of the DNA-damage response, including breakdown of the cell-cycle checkpoint mechanism. Although malfunction of cell-cycle checkpoints could be due to mishaps at any step in DNA repair, it is a strong indicator of a failure to properly process DNA double-strand breaks (DSBs). Indeed, the authors show that VPA-treated cells could not correctly repair such breaks.
Several proteins are responsible for repairing DSBs3, among them Sae2, Exo1, the MRX/N complex, Sgs1 (BLM) and Dna2. Robert et al. report a significant reduction in the association of Sae2 and Exo1 with broken DNA ends in VPA-treated cells. More intriguingly, the cellular levels of the two proteins were severely reduced in these cells, which would explain the failure of DSB repair. But why would VPA treatment affect Sae2 and Exo1 levels?
An analysis4 of all cellular proteins modified by acetylation of their lysine amino-acid residues identified some that mediate DNA repair, including Exo1. Although Sae2 was not among the acetylated proteins identified4, the study did point to the possibility that Sae2 is acetylated, and that its acetylated version is unstable. Deacetylation by HDACs would convert Sae2 to a stable form, but VPA treatment inhibits this.
Indeed, Robert and co-workers1 also find that Sae2 can be acetylated and that two HDACs — Hda1 and Rpd3, which have similar functions — promote its deacetylation. What's more, depletion of these two HDACs had similar effects to VPA treatment, leading to negligible Sae2 levels, heightened DNA-damage sensitivity and failure to activate cell-cycle checkpoints. Consistent with these observations, a recent study5 showed that the lysine deacetylase enzyme SIRT6 positively regulates the repair of DSBs through deacetylation of CtIP, the mammalian form of Sae2.
But this is only half the story. Why would inhibition of deacetylation (in other words, acetylation) destabilize Sae2? Considering the consequences of treating mammalian cells with VPA and other HDAC inhibitors2,6,7, Robert and colleagues propose — and provide evidence in yeast — that VPA stimulates autophagy, a degradation process that is normally linked to the cellular response to starvation and to organelle turnover. The authors further show that mutant cells defective in autophagy could overcome the inability of HDAC mutants to repair DSBs, presumably by stabilizing Sae2. They also saw the same spectrum of traits in a histone acetyltransferase mutant, which could not acetylate Sae2.
These findings are particularly noteworthy because they bring autophagy into the DNA-repair network. Autophagy can be triggered by inhibition of the enzyme TOR1 kinase8. Indeed, Robert et al. find that inhibiting this enzyme with the drug rapamycin induces autophagy and results in destabilization of Sae2.
From their results, the authors propose the following model (Fig. 1). Severely damaged DNA — such as DSBs that are difficult to repair — might become sequestered at the nuclear pores9, possibly to keep the cell's repair enzymes away from the bulk DNA that, although undamaged, could contain structures such as DNA nicks and gaps as a normal consequence of DNA replication. If the repair enzymes were close by, they could mistake these structures for damage and 'repair' them, leading to mutations and rearrangements.
In the case of Sae2, for example, although HDACs maintain it in an active state for repairing damaged DNA, at some point it will become acetylated and be expelled through the nuclear pore to the vacuole — the cellular site of autophagic degradation in yeast. Thus, a potentially dangerous enzyme is eliminated, preventing it from unnecessarily repairing replication-associated structures such as stalled replication forks. It remains unclear whether Sae2 is targeted for degradation after the completion of DNA repair and whether a checkpoint signal is involved.
Many further questions remain. Do other post-translational modifications, such as ubiquitination, phosphorylation and SUMOylation, regulate Sae2? Because of its association with autophagy, ubiquitination is most likely to play a part. But how does this relate to acetylation?
Also, several proteins that mediate the DNA-damage response — including Cdk1, Ku, MRN, Blm and Rfa1 — are acetylated1,4. Which of these are controlled through acetylation and autophagic degradation? Are nuclear pores an essential component of acetylation-promoted autophagy of DNA-repair complexes? Does a similar regulatory process occur during meiotic cell division, when DSBs abound and their repair is carefully choreographed? Is there an intranuclear cycle of acetylation–deacetylation, or is all acetylated Sae2 targeted for degradation?
These questions must be tackled before researchers can embark on exploring how this newly identified layer of DNA-damage regulation can be exploited to find targets for cancer therapy — a setting in which cells experience more DNA damage than usual.
Robert, T. et al. Nature 471, 74–79 (2011).
Göttlicher, M. et al. EMBO J. 20, 6969–6978 (2001).
Mimitou, E. P. & Symington, L. S. DNA Repair doi:10.1016/j.dnarep.2010.12.004 (2011).
Choudhary, C. et al. Science 325, 834–840 (2009).
Kaidi, A., Weinert, B. T., Choudhary, C. & Jackson, S. P. Science 329, 1348–1353 (2010).
Shao, Y., Gao, Z., Marks, P. A. & Jiang, X. Proc. Natl Acad. Sci. USA 101, 18030–18035 (2004).
Fu, J., Shao, C.-J., Chen, F.-R., Ng, H.-K. & Chen, Z.-P. Neuro-Oncology 12, 328–340 (2010).
Degenhardt, K. et al. Cancer Cell 10, 51–64 (2006).
Nagai, S. et al. Science 322, 597–602 (2008).
About this article
DNA Repair (2020)
A clinical trial with valproic acid and hydralazine in combination with gemcitabine and cisplatin followed by doxorubicin and dacarbazine for advanced hepatocellular carcinoma
Asia-Pacific Journal of Clinical Oncology (2020)
Aberrant BLM cytoplasmic expression associates with DNA damage stress and hypersensitivity to DNA-damaging agents in colorectal cancer
Journal of Gastroenterology (2017)
Proceedings of the National Academy of Sciences (2013)