Two papers last week addressed related puzzles in Huntington’s and related diseases, that is, why children of affected parents tend to get the disease at earlier ages than their parent, and how to possibly dampen the toxicity of the offending proteins. The number of neurodegenerative diseases caused by the appearance of abnormal glutamine repeats in different genes exceeds 30 now, with Huntington’s the most common one. Expansion of the CAG trinucleotide in protein coding regions leads to polyQ-expanded proteins with dominant toxic effects, and the longer the insertion, the earlier disease begins and the more severe its course. Strangely, the inserts tend to lengthen in successive generations, ever worsening the disease in a given family. Why this is so has been an intractable question, in part because the commonly used animal models of the disease do not replicate this genetic instability.

That has now changed, as Joonil Jung and Nancy Bonini of the University of Pennsylvania in Philadelphia report the successful induction of human-like repeat instability in a fruit fly model of polyQ disease. In the 1 March Science, they demonstrate that transcription of genes in germ line cells is behind the changes in repeat length. Using flies to probe the mechanism of the changes (which consisted mostly of expansions), they find that inhibiting DNA repair systems and, in particular, the CREB-binding protein (CBP), increases repeat instability. The results lead to the surprising conclusion that some toxic polyQ proteins, including huntingtin, might themselves promote repeat expansion through their well-established ability to sequester CBP and inhibit its activity.

This apparently circular logic of expansion suggests that treatments aimed at reducing polyQ protein toxicity could have the fringe benefit of preventing further expansion, as well. That would help affected families, who often have to watch their children and grandchildren develop symptoms at successively earlier ages. Besides the inherited expansion, this change can also occur during a person’s lifetime in somatic brain cells, possibly through similar mechanisms (Kennedy et al., 2003). One way to reduce polyQ toxicity would be to promote degradation of the offending proteins, and the second new paper, from Ole Isacson and colleagues at the McLean Hospital in Belmont, Massachusetts, shows that boosting the activity of the ubiquitin-dependent proteasome may offer a way to do that.

But first, the flies. It was in them that other scientists had previously demonstrated the link between CBP and the toxicity of mutant huntingtin protein (see ARF related news story). They showed that huntingtin inhibits CBP’s histone acetyltransferase activity, possibly explaining why the protein causes derangements in gene expression. Later work established that lowering CBP enhanced polyQ toxicity (Taylor et al., 2003). These studies have led to a promising therapeutic approach: Histone deacetylase inhibitors (which boost overall levels of acetylation) proved to suppress toxicity of polyQ proteins in fly or mouse models and are now in clinical development for HD (see review by Butler and Bates, 2006).

Jung and Bonini’s present work demonstrates that repeat expansion is part of the same story. In their work, flies with lower CBP levels showed decreased repeat stability of a polyQ-expanded ataxin transgene, whereas CBP overexpression stabilized the repeats. And treating the flies with the deacetylase inhibitor trichostatin A shut off the instability while restoring histone acetylation levels. For their mechanistic studies, Jung and Bonini used their fly model for the human disease spinocerebellar ataxia type 3, which expresses a mutated ataxin protein. However, the results are likely applicable to other polyQ diseases. The authors note that most triplet repeat disease genes are expressed in germ cells, and many polyQ proteins interact with CBP or other histone acetyl transferases. For example, flies with polyQ-expanded huntingtin also showed increased instability that was linked to germ line transcription or reduced CBP levels. Likewise, in a fragile X model, where the expansion occurs in noncoding regions, germ line transcription enhanced instability.

“Trinucleotide repeat instability has been viewed largely as a matter of DNA metabolism; however, our data suggest that repeat instability may be influenced by aspects of polyQ protein toxicity. Thus, treatments to curb polyQ protein pathology may also be effective means to help clamp repeated instability,” the authors conclude.

Enter the proteasome work from Isacson’s lab. It follows a previous study showing that the activity of the ubiquitin-dependent proteasome is reduced in cells from HD patients (Seo et al., 2004). In the February 28 PLoS ONE, first author Hyemyung Seo demonstrates that overexpressing a proteasome activator subunit (PA28γ) in fibroblasts from HD patients reverses compromised proteasome function. To test the strategy in a cell type affected by HD, they moved to rat striatal neurons expressing mutant huntingtin. The same proteasome activator protected those cells from neurotoxic stressors, including quinolinic acid or a proteasome inhibitor. “Although proteasome dysfunction is probably only one of multiple factors involved in the dynamic and progressive disease process of HD, our data at least demonstrate that proteasome actuators are relevant candidates for future comprehensive and effective treatment approaches to HD,” the authors write.—Pat McCaffrey


  1. The idea that proteasome dysfunction occurs in several neurodegenerative disorders has received quite a lot of attention in recent years. The strength of evidence varies between diseases, but the evidence in Huntington disease (HD) is reasonably strong. For example, genes for proteasome components modulate huntingtin toxicity in yeast but do not affect the toxicity of α-synuclein (Willingham et al., 2003). However, translating this to mammalian systems, and, hence, addressing whether proteasome dysfunction is an important part of the disease process in HD, has been less successful. For example, there are negative results reported, such as the lack of effect of the proteasome activator PA28γ in the R6/2 HD mice (Bett et al., 2006). In this study, Seo et al. report positive effects of increasing expression of PA28γ on cell viability in HD models. This is potentially exciting as it supports the idea that the proteasome might be a therapeutic target for HD, and perhaps for other neurodegenerative disorders. But, why is the evidence for proteasome involvement strong in cells (and in yeast) but weaker in the R6/2 mice?

    One reason Seo et al. discuss is that the R6/2 mice differ from end-stage HD in that the former have increased proteasome activity compared to decreased activity in the latter. This suggests that the model system might not go far enough in modeling the disease, especially its later stages. Of course, the later stages include dramatic cell loss, which might also be a confound for proteasome measurements, but the same investigators have shown previously that the proteasome deficit is seen in early disease stages and outside of the brain (Seo et al., 2004). So to test the hypothesis further, such as in vivo, one would need a model with decreased proteasome function where PA28γ expression could be modulated, something that is apparently true in YAC HD mice.

    Therefore, it might be feasible to establish whether the proteasome plays a role in HD and support the hypothesis that there is the possibility for disease modification. But, to my mind, this still leaves somewhat open the question of when proteasome inhibition contributes to HD (and even more so in other neurodegenerative disorders). The R6/2 mice, with no proteasome impairment, still have neurological phenotypes suggesting that proteasome inhibition is not a necessary component of disease. Perhaps the proteasome dysfunction is a later event in the disease course. Therefore, establishing whether PA28γ can modulate progression at different time courses in the disease will be an interesting and important question for the future.


    . Yeast genes that enhance the toxicity of a mutant huntingtin fragment or alpha-synuclein. Science. 2003 Dec 5;302(5651):1769-72. PubMed.

    . Proteasome impairment does not contribute to pathogenesis in R6/2 Huntington's disease mice: exclusion of proteasome activator REGgamma as a therapeutic target. Hum Mol Genet. 2006 Jan 1;15(1):33-44. PubMed.

    . Generalized brain and skin proteasome inhibition in Huntington's disease. Ann Neurol. 2004 Sep;56(3):319-28. PubMed.

    View all comments by Mark Cookson

Make a Comment

To make a comment you must login or register.


News Citations

  1. Drugs Slow Neurodegeneration in Fly Model of Huntington's

Paper Citations

  1. . Dramatic tissue-specific mutation length increases are an early molecular event in Huntington disease pathogenesis. Hum Mol Genet. 2003 Dec 15;12(24):3359-67. PubMed.
  2. . Aberrant histone acetylation, altered transcription, and retinal degeneration in a Drosophila model of polyglutamine disease are rescued by CREB-binding protein. Genes Dev. 2003 Jun 15;17(12):1463-8. PubMed.
  3. . Histone deacetylase inhibitors as therapeutics for polyglutamine disorders. Nat Rev Neurosci. 2006 Oct;7(10):784-96. PubMed.
  4. . Generalized brain and skin proteasome inhibition in Huntington's disease. Ann Neurol. 2004 Sep;56(3):319-28. PubMed.

Further Reading

Primary Papers

  1. . CREB-binding protein modulates repeat instability in a Drosophila model for polyQ disease. Science. 2007 Mar 30;315(5820):1857-9. PubMed.
  2. . Proteasome activator enhances survival of Huntington's disease neuronal model cells. PLoS One. 2007;2(2):e238. PubMed.