4 March 2005. Most of the neurons you have at birth will still be firing well into old age. If they weren’t, you might have trouble remembering your first love, or worse, your last. But does this stability come at a price? Because neurons don’t divide, they lose out on the opportunity for replication-based DNA repair, a quality control system that ensures near-perfect genome copies are passed on to the next generation of cells. In fact, recent evidence has fingered DNA damage as a factor in neurodegeneration (see related ARF live discussion) and suggests that past age 40, accumulated damage is sufficient to hamper gene expression (see ARF related news story). However, while DNA mutations are known to cause neurodegenerative diseases such as Alzheimer disease (AD), evidence linking a different kind of damage, DNA strand breaks, to neurodegenerative disease has been lacking—until now.
In this week's Nature, Keith Caldecott and colleagues at the University of Sussex in England, together with international collaborators in the USA, Canada, Sweden, and elsewhere in the UK, report that defective repair of single strand breaks in DNA causes spinocerebellar ataxia with axonal neuropathy-1, or SCAN1. This disease is caused by mutation in tyrosyl phosphodiesterase 1 (TDP1), a protein that is involved in repairing double-strand breaks that occur during replication of DNA.
“But wait—neurons don’t divide,” you say. And therein lies the crux of the matter. How can mutations in a protein that shouldn’t even be needed in neurons cause neurodegeneration?
To answer this, first author Sherif El-Khamisy and colleagues tested the ability of SCAN1 lymphoblastoid cells to cope with single-strand DNA breaks, which can occur independently of replication. El-Khamisy and colleagues treated these cells with camptothecin, a chemical that interferes with topoisomerase-mediated unraveling of DNA. Topoisomerase unties knots that appear in the DNA, but to do so it must break and re-ligate strands. Slowing down this process increases the likelihood that the breaks will not get fixed.
The authors found that, relative to normal cells, SCAN1 cells had about an eightfold higher incidence of DNA breaks when treated with camptothecin than did normal cells. In addition, these breaks remained even after the drug was removed. In contrast, the eightfold increase in breaks induced by upping the dose of the drug in normal cells was rapidly reversed once the drug was removed. The results indicate that TDP1 is required for fixing strand breaks.
To mimic the situation found in neurons, the authors treated the lymphoblasts with mimosine, which inhibits DNA replication. This completely eliminated breaks in normal cells but only reduced it by half in SCAN1 cells, suggesting that these breaks can indeed arise independently of replication. When the authors examined the breaks, they found that over 98 percent of them were of the single-strand variety, indicating that TDP1 is necessary for repairing single-strand breaks that occur independently of replication.
The authors propose that human cells possess a type of TDP1 that rapidly repairs topoisomerase-generated single-strand breaks. In support of this, they found that cells lacking an XRCC1-DNA ligase IIIα dimer, two proteins involved in single-strand break repair (SSBR), also accumulate breaks when treated with camptothecin. The authors also found that TDP1 binds to the ligase, suggesting that it may be part of the SSBR complex.
So how might TDP1 mutations cause SCAN1, a hereditary disease that usually strikes during the teenage years? One possibility is that the teenagers can’t repair oxidative damage. In fact, El-Khamisy and colleagues found that SCAN1 cells were less adept at fixing DNA breaks induced by hydrogen peroxide. Thirty minutes after the oxidant was removed, breaks in normal cells were almost fully repaired, while breaks in cells harboring the TDP1 mutation were about fivefold higher.
Though SCAN1 is not caused by a polyglutamine expansion, many other forms of neurodegenerative diseases are, including other spinocerebellar ataxias and Huntington disease. It is intriguing, therefore, that DNA strand breaks have been suggested as playing an important role in the expansion of the CAG trinucleotide repeats that code for polyglutamine (see ARF related news story). The implication that many of these diseases ultimately derive from a failure to repair DNA will be worth investigating.—Tom Fagan.
El-Khamisy SF, Saifi GM, Weinfeld M, Johansson F, Helleday T, Lupski JR, Caldecott KW. Defective DNA single-strand break repair in spinocerebellar ataxia with axonal neuropathy-1. Nature 2005 March 3;434:108-113.