Huntington disease (HD) is a neurodegenerative disorder caused by the repetition of a cytosine-adenine-guanine (CAG) trinucleotide coding for the amino acid glutamine. The repeat occurs in the huntingtin gene and results in an expanded huntingtin (Htt) protein containing a polyglutamine (polyQ) tract. Though expanded huntingtin accumulates in neurons, there is continuing debate over whether these intracellular inclusions are damaging or protective (see ARF related news story), and as yet no consensus has emerged as to how mutant Htt causes neurodegeneration. Nonetheless, three recent papers show how RNAi, resveratrol, and inhibitors of kyneurenine-3-monooxygenase may offer hope as therapeutics for HD and other polyglutamine disorders.

Reporting in the April Nature Genetics, Christian Neri and colleagues at INSERM in Paris, France, show that resveratrol can attenuate polyQ huntingtin-mediated toxicity in worms and mammalian neurons. The polyphenol, found in grapes and wine, has gotten a lot of press since researchers found it dramatically increases lifespan in yeast, worms, and flies (see, for example, Howitz et al., 2003 and that it can protect axons against degeneration (see ARF related news story). These actions are attributed to activation of sirtuins, a family of NAD-dependent histone deacetylases that includes SIRT1, which may extend the life of mammalian cells (see ARF related news story).

First author Alex Parker and colleagues found that resveratrol improved HD-like pathology in the roundworm Caenorhabditis elegans. Though some would call these creatures primitive, they do have motor neuron function of a sort, which the authors tapped into by measuring the response of the animals’ tails to mechanical stimuli. Parker found that tail mechanosensitivity jumped 20 percent in worms fed resveratrol. Htt aggregation was unchanged, however, which supports suggestions that aggregates per se are not the bad guys in HD. Worms given another sirtuin activator, fisetin, were similarly feisty, as were those overexpressing sir-2.1, the worm homolog of SIRT1. The sirtuin activators did not work in worms expressing mutated, loss-of-function sir-2.1.

Worms, of course, may not represent a great model for human disease, but Parker and colleagues obtained similar results when they used striatal cells from mice that express Hdh109Q, the expanded mouse homolog of the human gene. In these cells, resveratrol rescued dystrophic processes by about 35 percent, again, only if active sir-2 was present. The polyphenol also reduced cell mortality by about 40 percent, but had no effect on Hdh expression or aggregation.

The effect of resveratrol on these HD models might be related to apoptosis. Sirtuins are thought to attenuate apoptotic signals through their actions on the proapoptotic transcription factors forkhead (see Motta et al., 2004 and Bax (see ARF related news story), and Parker found that loss-of-function mutations in forkhead family member daf16 also rescued tail sensitivity in worms by about 15 percent. Sirtuins are also activated when animals are placed on a caloric restriction diet. In this regard, it is telling that a loss-of-function mutation in the gene age-1, coding for a kinase that mediates insulin-like signaling, also rescued tail sensitivity by about 10 percent. Though it is unclear whether resveratrol’s effects are mediated by any of these signaling pathways, the authors conclude that sirtuin activators “may be useful in the development of therapeutic strategies for Huntington disease.”

However, things are never simple. The second Nature Genetics paper, this one from Paul Muchowski and colleagues at the University of Seattle, Washington, and the University of Maryland School of Medicine, Baltimore, suggests that activating histone deacetylases might have the opposite effect, exacerbating polyQ toxicity. First author Flaviano Giorgini and colleagues conducted screens of over 4,000 strains of yeast for gene deletions that suppress huntingtin toxicity. They found that absence of 28 proteins makes the yeast grow stronger. Twenty-four of these proteins were of known function and two, Ume1 and Rxt3, are members of the yeast Rpd3 histone deacetylase complex. Because Ume1 is required for full activity of the complex, this data indicates histone deacetylases may actually contribute to huntingtin toxicity, a suggestion that has much support. For example, inhibitors of histone deacetylases slow neurodegeneration in a fly model of Huntington disease (see ARF related news story), while lack of histone acetylases has indirectly linked deacetylation to huntingtin toxicity (see ARF related news story).

So how can both activation and inhibition of histone deacetylases protect against huntingtin toxicity? Perhaps the answer lies in substrate specificity. It is possible that NAD-dependent and NAD-independent deacetylases can have opposite effects on polyQ pathology, for example. But what is also intriguing is that Muchowski’s group found that deletion of kyneurenine-3-monooxygenase (KMO), an enzyme involved in NAD synthesis, strongly suppressed Htt128Q toxicity in yeast. Does this suggest that the NAD-dependent sirtuins also contribute to huntingtin toxicity, which would be at total odds with the data from Neri’s group? Perhaps not. Because there are alternative pathways for NAD synthesis, it is unclear if sirtuin activity is affected in KMO knockout mice. Instead, Muchowski and colleagues suggest a different reason for the increased Htt toxicity found in yeast with KMO—oxidative damage.

Two intermediaries, quinolinic acid and 3-hydroxykynurenine (3HK), which lie downstream of KMO in the NAD synthesis pathway, produce reactive oxygen species (ROS). They are also known to cause neurotoxicity in animals and are elevated in early-stage Huntington disease (see Guidetti et al., 2000 and Guidetti et al., 2004). Giorgini and colleagues found that both compounds are absent from KMO-negative yeast, and that that ROS levels were normal in these strains, even when they expressed Htt128Q. In contrast, the authors found that wild-type yeast expressing the expanded huntingtin had ROS levels eightfold higher than controls. But, when Girogini added the KMO inhibitor Ro 61-8048 to these sickly yeast, ROS were reduced by about twofold and growth was partially rescued. The author found these results so promising that “…preclinical trials with the compound are currently underway using mouse models of HD,” they write.

An alternative to tackling the toxicity of expanded huntingtin is to target production of the protein. In this week’s PNAS, Beverly Davidson and colleagues at the University of Iowa, Iowa City, and the NIH at Bethesda, Maryland, report that they have used RNA interference (RNAi) to ablate huntingtin mRNA in a mouse model of HD.

First author Scott Harper and colleagues used adenoassociated viral particles to infect the brains of transgenic mice (HD-N171-82Q) with an expression system that produces short hairpin RNAs (shRNAs) that target exon 2 of the huntingtin transcript (see ARF related news story). About three weeks after infection, Harper and colleagues detected robust levels of 50- and 21-nucleotide-long shRNAs in the brains of treated mice. They also found that the shRNA reduced expression of mutant Htt by over twofold, and almost completely eliminated Htt inclusions in cells where Htt expression was silenced—the system used to drive shRNA production also expresses enhanced green fluorescent protein, enabling the authors to see exactly which cells produce the interfering RNA.

But it was in behavioral tests where the system came through with flying colors. Huntington disease affects the motor neurons, leading to involuntary movements, balance problems, and muscle weakness. Transgenic mice expressing the mutant huntingtin are no different; they, too, have trouble walking and maintaining balance. But Harper found that animals treated with RNAi had dramatically improved performance. On the rotarod, for example, which is the mouse equivalent of the gymnast’s balance beam, treated animals learned to hang on for almost 350 seconds; 250 seconds was all that untreated mice could muster. Harper and colleagues also found that stride length was significantly improved in the treated animals, and the researchers concluded that the “data suggest the feasibility of treating HD by directly reducing mutant Htt gene expression by using RNAi and support its general applicability to treating other dominant neurodegenerative disorders.”—Tom Fagan


  1. RNAi Therapy Works in Animal Models

    Scott Harper et al. recently demonstrated that RNA interference (RNAi) can treat Huntington disease in an animal model (Harper et al., 2005). This work, together with a previous published experiment from the same group on treatment of spinal cerebellar ataxia (Xia et al., 2004), and two other experiments on treatment of ALS (Ralph et al., 2005; Raoul et al., 2005), demonstrates the concept of RNAi therapy for neurodegenerative diseases.

    The common approach in these experiments was to deliver RNAi using viral vectors. All showed in vivo knockdown of the target gene and phenotypic improvement. These are very encouraging developments that bring RNAi one step closer to clinical application. Here I provide some background about these experiments and discuss some challenges that we still need to meet in order to realize the full therapeutic potential of RNAi.

    In general, genetic disorders can be caused by two types of genetic mutations. One causes the gene to lose its function and the other causes the gene to gain a function—the gain could be either a novel function or an enhancement of an existing function. The treatment strategy for the loss-of-function type of mutations is to deliver to the patient a normal copy of the mutated gene so that its function can be replaced. But for a long time the conceptual framework for treatment of gain-of-function mutations was less obvious. However, RNAi has provided a conceptual basis for this kind of treatment (see ARF Live Discussion).

    RNAi is a cellular function that is widely conserved in eukaryotes. Triggered by double-stranded RNA (dsRNA), RNAi destroys the target RNA that shares sequence homology with the dsRNA. The mechanism of RNAi is not understood completely, but a key executor is a protein complex called RNA-induced silencing complex, or RISC, which contains proteins and a single-stranded RNA (guide strand) derived from the dsRNA that the cells are originally exposed to. This guide strand is capable of recognizing the target RNA by Watson-Crick base pairing. After the pairing, RISC cleaves the target RNA. Thus, RNAi can destroy the target RNA specifically. When the target is an mRNA, RNAi can selectively knock down the expression of the target protein (Tomari and Zamore, 2005). This makes RNAi ideal for treatment of diseases caused by gain-of-function type of genetic mutations (see ARF Live Discussion).

    A serious challenge to the success of RNAi therapy is the delivery of RNAi in vivo. Authors of the above four papers met this challenge by demonstrating that viral vector-delivered RNAi can treat neurodegenerative diseases in animal models. These experiments used viral vectors that express short hairpin RNA (shRNA) that can be processed to short RNA duplexes called small interfering RNA (siRNA) (Shi, 2003). The siRNAs mediated RNAi against mutant ataxin-1, SOD1, and huntingtin mRNAs, which cause spinal cerebellar ataxia 1 (SCA-1), amyotrophic lateral sclerosis (ALS) and Huntington disease, respectively. These successes are not surprising since all of these diseases are caused by gain-of-function gene mutations. Although the nature of the gained function is not clear, it is clear that disease severity correlates with the levels of mutant protein. Therefore, lowering the mutant protein levels will protect neurons from the toxic effects of these proteins.

    What are the future challenges in developing RNAi therapy? Here I raise a few: First, how long is shRNA expressed after the viral delivery and to what extent and for how long is the target gene silenced? These questions have not been answered in these experiments. Therefore, it is not clear whether the observed therapeutic effects are a result of transient silencing or a sustained silencing of the target genes. A related question is, what is the best time to administer the RNAi therapy, before or after the disease onset? In all experiments the RNAi therapy was administered before the disease onset. It will be important to determine whether, by administration of RNAi therapy after the onset of the disease, the disease can be reversed.

    Next, as a general rule, is it necessary to administer mutant-allele-specific RNAi? In these four cases the siRNAs were not specifically designed to silence the mutant gene because the mutant genes are human transgenes that differ in sequences from the mouse endogenous genes. Silencing the transgenes does not affect expression of endogenous mouse proteins. In human patients, however, shRNAs would silence both the mutant and the wild-type gene expression. This may be problematic because the wild-type functions of these genes are necessary for normal cellular function, and animal experiments show that in some cases they are essential (Cattaneo et al., 2001; Ding et al., 2003). In humans, the phenotypes for lack of these genes are not known, but they could be serious, given the animal data. To overcome this problem, allele-specific silencing or RNAi with replacement gene therapy may be necessary (Ding et al., 2003; Miller et al., 2003; Xia et al., 2005).

    Third, are the viral vectors the best way for delivery of RNAi therapy? Delivery using viral vectors carries risks associated with gene therapy (Kimmelman, 2005). An alternative is to deliver siRNA directly to patients. A recent study has demonstrated that chemically modified siRNAs can be systemically administered and can knock down specific genes in multiple tissues (Soutschek et al., 2004; see ARF related news story). However, siRNAs with chemical modifications that allow the siRNA to penetrate the blood-brain barrier have not been reported. Nevertheless, siRNA or chemically modified siRNA may be directly delivered to the CNS to induce specific knockdown (Thakker et al., 2004). Direct delivery to the CNS does carry disadvantages, such as risks and costs associated with surgery and constant consumption of drugs. Long-term risks of toxicity with chronic administration of siRNA or chemically modified siRNA are yet to be determined.

    In conclusion, the new results are exciting, and they are the first steps toward developing RNAi therapy. There is little doubt that most, if not all, mammalian cells possess the RNAi mechanism. The key to therapy rests on the development of highly efficient and safe delivery methods. If this can be achieved, RNAi can not only be used to treat genetic disorders that are caused by dominant, gain-of-function mutations, but also sporadic diseases where the disease pathways are understood. For example, if the amyloid hypothesis for Alzheimer disease is correct (Hardy and Selkoe, 2002), RNAi can be used to downregulate BACE and/or other proteins that are involved in the production of Aβ, a substrate for β amyloid that is postulated to cause neuronal degeneration.


    . Loss of normal huntingtin function: new developments in Huntington's disease research. Trends Neurosci. 2001 Mar;24(3):182-8. PubMed.

    . Selective silencing by RNAi of a dominant allele that causes amyotrophic lateral sclerosis. Aging Cell. 2003 Aug;2(4):209-17. PubMed.

    . The amyloid hypothesis of Alzheimer's disease: progress and problems on the road to therapeutics. Science. 2002 Jul 19;297(5580):353-6. PubMed.

    . RNA interference improves motor and neuropathological abnormalities in a Huntington's disease mouse model. Proc Natl Acad Sci U S A. 2005 Apr 19;102(16):5820-5. PubMed.

    . Recent developments in gene transfer: risk and ethics. BMJ. 2005 Jan 8;330(7482):79-82. PubMed.

    . Allele-specific silencing of dominant disease genes. Proc Natl Acad Sci U S A. 2003 Jun 10;100(12):7195-200. PubMed.

    . Silencing mutant SOD1 using RNAi protects against neurodegeneration and extends survival in an ALS model. Nat Med. 2005 Apr;11(4):429-33. PubMed.

    . Lentiviral-mediated silencing of SOD1 through RNA interference retards disease onset and progression in a mouse model of ALS. Nat Med. 2005 Apr;11(4):423-8. PubMed.

    . Mammalian RNAi for the masses. Trends Genet. 2003 Jan;19(1):9-12. PubMed.

    . Therapeutic silencing of an endogenous gene by systemic administration of modified siRNAs. Nature. 2004 Nov 11;432(7014):173-8. PubMed.

    . Neurochemical and behavioral consequences of widespread gene knockdown in the adult mouse brain by using nonviral RNA interference. Proc Natl Acad Sci U S A. 2004 Dec 7;101(49):17270-5. PubMed.

    . Perspective: machines for RNAi. Genes Dev. 2005 Mar 1;19(5):517-29. PubMed.

    . RNAi suppresses polyglutamine-induced neurodegeneration in a model of spinocerebellar ataxia. Nat Med. 2004 Aug;10(8):816-20. PubMed.

    . An RNAi strategy for treatment of amyotrophic lateral sclerosis caused by mutant Cu,Zn superoxide dismutase. J Neurochem. 2005 Jan;92(2):362-7. PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. New Microscope Resolves Role of Huntington Inclusions—Neuroprotection
  2. The Wine and Wherefore of Wallerian Degeneration
  3. Who Says Chivalry is Dead?—Sir2 Fights Against Aging in Mammals
  4. Drugs Slow Neurodegeneration in Fly Model of Huntington's
  5. Modeling Polyglutamine Diseases in Yeast Provides Support for Histone Deacetylase Connection
  6. RNA Interference in Vivo—Viral Vectors Deliver

Paper Citations

  1. . Small molecule activators of sirtuins extend Saccharomyces cerevisiae lifespan. Nature. 2003 Sep 11;425(6954):191-6. PubMed.
  2. . Mammalian SIRT1 represses forkhead transcription factors. Cell. 2004 Feb 20;116(4):551-63. PubMed.
  3. . Early kynurenergic impairment in Huntington's disease and in a transgenic animal model. Neurosci Lett. 2000 Apr 14;283(3):233-5. PubMed.
  4. . Neostriatal and cortical quinolinate levels are increased in early grade Huntington's disease. Neurobiol Dis. 2004 Dec;17(3):455-61. PubMed.

Further Reading

Primary Papers

  1. . Resveratrol rescues mutant polyglutamine cytotoxicity in nematode and mammalian neurons. Nat Genet. 2005 Apr;37(4):349-50. PubMed.
  2. . A genomic screen in yeast implicates kynurenine 3-monooxygenase as a therapeutic target for Huntington disease. Nat Genet. 2005 May;37(5):526-31. PubMed.
  3. . RNA interference improves motor and neuropathological abnormalities in a Huntington's disease mouse model. Proc Natl Acad Sci U S A. 2005 Apr 19;102(16):5820-5. PubMed.