A recent study tied germ line repeat expansion to gene transcription and associated DNA repair pathways for one polyQ protein (see ARF related news story). This phenomenon leads to worsening disease in successive generations of affected families. Now, Cynthia McMurray and colleagues from the Mayo Clinic in Rochester, Minnesota, with collaborators at the University of Oslo in Norway and the NIH, show that expansion in somatic cells also results from DNA repair mechanisms, this time in response to oxidative damage. Because somatic expansions, which they show occur in postmitotic neurons, increase the toxicity of the Huntington protein, they could affect when the disease begins in a given person, and the severity of symptoms. The results provide a direct mechanistic link between age-related increases in oxidative DNA damage and neuronal toxicity.

The new paper, published in Nature online April 22, fleshes out a story McMurray presented in preliminary form in 2004 at a protein misfolding conference in San Diego (see ARF related news story). The tale starts with the observation that in transgenic mice bearing a piece of the Huntington gene with a pathologically high number of 118 CAG repeats, the repeat length begins to increase further in brain cells in middle age, and continues increasing until death. Reasoning that this phenomenon might explain the age-related onset of HD, lead author Irina Kovtun set out to determine how it occurred.

Her experiments showed that expansion with aging is tissue-specific, occurring in the liver and brain, but not the tail. Oxidative damage was implicated by the observation that CAG growth correlated with the extent of oxidized bases in DNA from the different tissues.

To ask if oxidative damage was able to induce CAG expansion, the researchers treated cultured human fibroblasts from Huntington patients with high concentrations of hydrogen peroxide to damage DNA. As a result, CAG triplets in the huntingtin gene expanded, whereas non-CAG repetitive elements elsewhere in the genome did not. In the fibroblasts, single-strand breaks appeared after peroxide treatment, a sign that the base excision repair machinery was working on the DNA to remove and replace oxidized bases. The breaks were mended within 2 hours, giving a window of opportunity for expansion.

But how could repair be responsible for CAG expansion? To test whether DNA repair enzymes are involved, the researchers crossed HD mice with animals deficient in 8-oxoG-DNA glycosylase (OGG1), which removes 8-oxo-guanine residues. On average, the crossed mice showed less age-dependent expansion, though the effect was not absolute. A third of the knockout mice still showed expansion with age comparable to mice expressing OGG1. Nonetheless, the suppression was surprising given that many types of DNA lesions develop in cells, not just 8-oxoG, and other DNA glycosylases can back up OGG1 and repair 8-oxoG lesions. However, knocking out either of two other glycosylases had no effect, leaving OGG1 as a major factor in age-dependent expansions in vivo. The investigators speculate that this could be so if OGG1 preferentially binds to CAG sequences, while the other enzymes do not.

Finally, the researchers used an in-vitro repair system to pin down OGG1’s action on synthetic DNA templates with an inserted 8-oxoG lesion. When they reconstituted the repair pathway, they found base excision by OGG1 proceeded apace on either random or CAG repeat templates. However, the gap-filling polymerase that came in afterwards generated longer products on the CAG template while generating correct repairs on the random sequence. The researchers noted a tendency to add trimers to the incorrectly mended sequence, mimicking expansion in vivo.

“Age-dependent CAG expansion provides a direct molecular link between oxidative damage and toxicity in postmitotic neurons through a DNA damage response, and error-prone repair of single strand breaks,” the authors conclude. HD requires an inherited expanded CAG tract. The authors propose that, on top of the initial inherited expansion, a “toxic oxidation cycle” occurs where somatic mutations, perhaps exacerbated by oxidative stress created by the presence of toxic proteins, accelerate disease and affect the time of onset or severity of symptoms, or both. Implicating OGG1 specifically in the polyQ repeat diseases could offer a new point for intervention. In addition, OGG1 has been implicated in other, non-polyQ neurodegenerative diseases. Recent work shows that in Parkinson disease, both 8-oxo-guanine and OGG1 activity are increased (Nakabeppu et al., 2007; Fukae et al., 2007). Oxidative damage to DNA is increased in Alzheimer disease, and a recent study reports mutations in the OGG1 gene that lower its activity in a subset of AD patients (Mao et al., 2007).—Pat McCaffrey.

Reference:
Kovtun IV, Liu Y, Bjoras M, Klungland A, Wilson SH, McMurray CT. OGG1 initiates age-dependent CAG trinucleotide expansion in somatic cells. Nature. 2007 Apr 22; [Epub ahead of print] Abstract