A safe fix for alcohol-derived DNA damage

A by-product of alcohol metabolism can damage the genome by crosslinking opposing DNA strands. The discovery of a safe mechanism that reverses such damage might open up avenues of research for drug discovery.
Irene Gallina is in the Faculty of Health and Medical Sciences, Novo Nordisk Foundation Center for Protein Research, University of Copenhagen, DK-2200 Copenhagen, Denmark.

Search for this author in:

Julien P. Duxin is in the Faculty of Health and Medical Sciences, Novo Nordisk Foundation Center for Protein Research, University of Copenhagen, DK-2200 Copenhagen, Denmark.

Search for this author in:

Aldehydes are highly reactive molecules that can enter the body from the environment, and can be produced by cellular metabolic processes. One aldehyde relevant to human health is acetaldehyde, which is produced when cells process ingested alcohol. If acetaldehyde accumulates in cells, it reacts with DNA and can link two strands together, generating an extremely harmful form of damage known as a DNA interstrand crosslink1 (ICL). ICLs are also produced by many anticancer drugs, to kill tumour cells. Writing in Nature, Hodskinson et al.2 report the discovery of a mechanism for repairing acetaldehyde-induced ICLs that is safer than the commonly used route.

An inability to repair ICLs is linked to the rare genetic disease Fanconi anaemia (FA). This condition is caused by mutations in any one of 22 FANC genes, which encode proteins that participate in ICL repair3. People who have FA experience genomic instability, bone-marrow failure and premature ageing, and have a high risk of developing cancer. Since the 1970s, scientists have known that the cells of people who have FA are exquisitely sensitive to ICL-inducing drugs4, but it was not until 2011 that researchers found genetic evidence5 suggesting that acetaldehyde-derived DNA damage is a driving force of FA. How this damage is repaired was unknown.

The need to clear acetaldehyde from cells to prevent DNA damage (lesions) became evident after the in vivo identification of a two-tier system in mice that protects against this highly reactive molecule5. The first tier of protection involves the enzyme aldehyde dehydrogenase 2 (ALDH2), which converts acetaldehyde to harmless acetate molecules (Fig. 1). Inactivation of this enzyme is common in members of Asian populations, and is associated with a higher incidence of alcohol-derived cancers6. The second tier is repair of the DNA damage generated by acetaldehyde.

Figure 1

Figure 1 | Cellular defences against acetaldehyde. When humans ingest alcohol, it is converted in the liver to toxic acetaldehyde by the enzyme alcohol dehydrogenase (ADH). Acetaldehyde can also be formed by other metabolic processes, or come from the environment (not shown). The compound is detoxified by another enzyme, aldehyde dehydrogenase 2 (ALDH2), but still sometimes accumulates in cells, in which it forms interstrand crosslinks (ICLs) between bases in DNA molecules. This damage can be repaired by the Fanconi anaemia pathway, in a process that involves the formation of DNA breaks either side of the ICL. However, DNA breaks are potentially dangerous, and can lead to harmful chromosome rearrangements and cancer. Hodskinson et al.2 report a second pathway for ICL repair in which the crosslink, rather than a DNA strand, is cut. This completely restores one of the bases that was crosslinked, and leaves an adduct on the other. This repair process prevents chromosomal rearrangements.

Because the combined inactivation of FANC and ALDH2 genes recapitulates the characteristics of FA in mice, it is suspected that ICLs are the cytotoxic (cell-killing) lesions generated by acetaldehyde5. Consistent with this view is the observation that FA severity correlates7 with the presence of an ALDH2 mutation in Japanese people who have FA. However, direct investigation of these crosslinks is not possible in cellular or in vivo systems using available technologies. Hence, whether acetaldehyde-induced ICLs accumulate in people who have FA remains a crucial unanswered question.

A previously reported cell-free in vitro system derived from frog eggs8 has been widely used to study the mechanisms underlying repair of ICLs induced by other agents, including the anticancer drug cisplatin9. This system allows DNA molecules that contain a single, site-specific DNA lesion to be analysed. In the case of cisplatin-induced ICLs, the cell-free system revealed a sophisticated repair mechanism that depends on FANC proteins9,10. This mode of repair requires DNA replication and cuts DNA strands to ‘unhook’ and remove the ICL (Fig. 1).

In their work, Hodskinson et al. undertook the enormous challenge of synthesizing a DNA molecule containing a single, site-specific acetaldehyde ICL, and then investigated how the lesion is repaired in the cell-free system. They found that this repair process requires an active FA pathway (a mechanism that involves FANC proteins). This is consistent with genetic evidence that FANC proteins are required in the two-tier system that protects against acetaldehyde damage. However, the authors unexpectedly discovered that about half of the crosslinks are fixed by a second, faster mechanism. Further investigation revealed that this second route also involves DNA replication, but is independent of the FA pathway.

Surprisingly, in the fast repair route, no cuts are made to the DNA strands; instead, the ICL is probably cut within the crosslink. This mode of repair results in the reversion of the crosslink to an undamaged base on one of the DNA strands, but leaves an adduct on the other strand (Fig. 1), which specialized DNA-replication enzymes can bypass to complete repair. This mechanism is reminiscent of the one that fixes ICLs generated by the drug psoralen11, but involves different enzymes. By avoiding DNA breaks — which are associated with genomic rearrangements, one of the hallmarks of cancer and ageing — the fast repair mechanism has an important advantage over the FA pathway. Taken together, Hodskinson and colleagues’ findings provide a holistic glimpse of how acetaldehyde-derived crosslinks are cleared from DNA, and support the idea that these lesions contribute to FA.

The authors do not identify a protein that cleaves the crosslinks in the newly described repair route. One can therefore only speculate as to whether cleavage occurs spontaneously as a consequence of mechanical forces generated during replication as the DNA unwinds, or is the result of enzymatic activity. If it is indeed an enzymatic process, identifying the components of the pathway will be a challenge, but could open up opportunities for therapies: stimulation of the pathway might alleviate the symptoms of FA, or reduce the incidence of alcohol-derived cancers.

The identification of the protein(s) involved in the crosslink cleavage would also allow in vivo experiments to test whether impairment of the alternative repair route increases acetaldehyde toxicity, especially under conditions in which this molecule is not detoxified by metabolism. Furthermore, mutations in the genes encoding proteins involved in this pathway might reveal the existence of a new group of people who have an FA-like disorder. In the meantime, Hodskinson and colleagues’ study underlines the need to develop better assays to study ICLs and other types of DNA damage in cells. By studying the repair of specific DNA lesions induced by compounds such as acetaldehyde, or other mutagens that arise in the body, we are likely to uncover other cellular defence mechanisms against cytotoxic DNA damage.

Nature 579, 499-500 (2020)

doi: 10.1038/d41586-020-00462-1


  1. 1.

    Wang, M. et al. Chem. Res. Toxicol. 13, 1149–1157 (2000).

  2. 2.

    Hodskinson, M. R. et al. Nature 579, 603–608 (2020).

  3. 3.

    Fiesco-Roa, M. O., Giri, N., McReynolds, L. J., Best, A. F. & Alter, B. P. Blood Rev. 37, 100589 (2019).

  4. 4.

    Sasaki, M. S. & Tonomura, A. Cancer Res. 33, 1829–1836 (1973).

  5. 5.

    Langevin, F., Crossan, G. P., Rosado, I. V., Arends, M. J. & Patel, K. J. Nature 475, 53–58 (2011).

  6. 6.

    Chang, J. S., Hsiao, J.-R. & Chen, C.-H. J. Biomed. Sci. 24, 19 (2017).

  7. 7.

    Hira, A. et al. Blood 122, 3206–3209 (2013)

  8. 8.

    Walter, J., Sun, L. & Newport, J. Mol. Cell 1, 519–529 (1998).

  9. 9.

    Räschle, M. et al. Cell 134, 969–980 (2008).

  10. 10.

    Knipscheer, P. et al. Science 326, 1698–1701 (2009).

  11. 11.

    Semlow, D. R., Zhang, J., Budzowska, M., Drohat, A. C. & Walter, J. C. Cell 167, 498–511 (2016).

Download references

Nature Briefing

An essential round-up of science news, opinion and analysis, delivered to your inbox every weekday.