DiagnosisClinical DiagnosisThe diagnosis of spinocerebellar ataxia type 3 (SCA3), also known as Machado-Joseph disease (MJD), is suggested in individuals with the following findings [Lima & Coutinho 1980, D’Abreu et al 2010]: Progressive cerebellar ataxia and pyramidal signs associated to a variable degree with a dystonic-rigid extrapyramidal syndrome or peripheral amyotrophy Minor (but more specific) clinical signs such as progressive external ophthalmoplegia, dystonia, action-induced facial and lingual fasciculation-like movements, and bulging eyes Family history consistent with autosomal dominant inheritance Because the clinical findings are shared with many other dominantly inherited ataxias, diagnosis of SCA3 rests upon molecular genetic testing. Molecular Genetic TestingGene. ATXN3 (known previously as MJD1) is the only gene in which mutations are known to cause SCA3. A polymorphic CAG repeat in ATXN3 is unstable and is expanded to an abnormal range in all individuals with SCA3. The CAG trinucleotide repeat in ATXN3 encodes a polyglutamine tract in the disease protein, ataxin-3 (see Table 2).Allele sizes. The following allele sizes are observed in SCA3 [Potter N et al, unpublished; the Ataxia Molecular Diagnostics Testing Group; compiled October 2002, updated August 2005]: Normal allele. Fewer than 44 CAG repeats [Padiath et al 2005]. Overall, 93.5% of normal alleles have fewer than 31 CAG repeats. Mutable normal allele (also called intermediate allele). Unknown. Mutable normal alleles have yet to be convincingly associated with a phenotype but can manifest meiotic instability, rarely resulting in a pathologic expansion in a subsequent generation [Maciel et al 2001]. Reduced penetrance allele. Between 45 to 51 CAG repeats [Gu et al 2004, Padiath et al 2005]. Individuals with a reduced penetrance allele may or may not manifest the disorder during their lifetime. Abnormal allele with full penetrance. Alleles with 52 to 86 CAG repeats are associated with the SCA3 phenotype. Clinical testing Targeted mutation analysis. Testing to determine the number of ATXN3 CAG repeats is performed typically by PCR amplification of the trinucleotide repeat region followed by gel or capillary electrophoresis. PCR analysis may be used to detect trinucleotide repeat expansions up to approximately 100 repeats; however, the upper limit of the size of the expansion detected may vary by laboratory. Table 1. Summary of Molecular Genetic Testing Used in Spinocerebellar Ataxia Type 3View in own windowGene SymbolTest MethodMutation Detected Mutation Detection Frequency by Test Method ^{1Test Availability ATXN3Targeted mutation analysisAbnormal number of CAG trinucleotide repeats100%Clinical 1. The ability of the test method used to detect a mutation that is present in the indicated geneInterpretation of test results. The presence of one disease-causing allele is diagnostic. Testing StrategyTo confirm/establish the diagnosis in a proband requires identification of an abnormally expanded ATXN3 CAG repeat. Predictive testing for at-risk asymptomatic adult family members requires prior identification of an abnormally expanded CAG repeat in ATXN3 in an affected family member. Prenatal diagnosis and preimplantation genetic diagnosis (PGD) for at-risk pregnancies require prior identification of an abnormally expanded CAG repeat in ATXN3 in an affected family member. Genetically Related (Allelic) DisordersNo other phenotypes are known to be associated with mutations in ATXN3.}

Natural HistorySpinocerebellar ataxia type 3 (SCA3), also known as Machado-Joseph disease (MJD), is characterized primarily by cerebellar ataxia and pyramidal signs variably associated with a dystonic-rigid extrapyramidal syndrome or peripheral amyotrophy.The age of onset of SCA3 is highly variable but most commonly in the second to fifth decade. In a large cohort of affected individuals from the Azores, the mean onset was age 37 years. The variable range of symptom onset largely reflects differences in the length of the CAG repeat.Presenting features include gait problems, speech difficulties, clumsiness, and often visual blurring and diplopia. Progressive ataxia, hyperreflexia, nystagmus, and dysarthria may occur early in the disease. Upper motor neuron signs often become prominent early on, and in some families may resemble hereditary spastic paraplegia [Wang et al 2009, Gan et al 2009]. Ambulation becomes increasingly difficult, leading to the need for assistive devices (including wheelchair) ten to 15 years following onset. Saccadic eye movements become slow and ophthalmoparesis develops, resulting initially in up-gaze restriction. Disconjugate eye movements result in diplopia. At the same time, a number of other "brain stem" signs develop, including temporal and facial atrophy, characteristic action-induced perioral twitches, vestibular symptoms, tongue atrophy and fasciculations, dysphagia, and poor ability to cough and clear secretions. Often, a staring appearance to the eyes is observed, but neither this nor the perioral fasciculations are specific for SCA3. Other findings may include the following:Vocal cord paralysis, described in three of 19 persons with SCA3 [Isozaki et al 2002] Vestibular dysfunction [Yoshizawa et al 2004] Autonomic problems, including bladder and thermoregulation disturbances [Yeh et al 2005, França et al 2010] A disabling sleep disturbance, rapid eye movement behavior disorder [Friedman 2002, Friedman et al 2003], and restless legs syndrome [Schols et al 1998, Van Alfen et al 2001, D’Abreu et al 2009, Pedroso et al 2011]. Chronic pain, often in the lumbosacral region, may precede onset of ataxia [França et al 2007]Individuals with SCA3 have been found to have impaired executive and emotional functioning that is unrelated to ataxia severity; however, these individuals do not meet the criteria for dementia [Zawacki et al 2002]. Verbal fluency and visual memory deficits have been noted [Kawai et al 2004].Later, evidence of a peripheral polyneuropathy [França et al 2009] may appear with loss of distal sensation, ankle reflexes, and sometimes other reflexes as well, and with some degree of muscle wasting. Severe ataxia of limbs and gait (with either hyperreflexia or areflexia) associated with muscle wasting is observed. Sitting posture is compromised, with affected individuals assuming various tilted positions.Late in the disease course, individuals are wheelchair bound and have severe dysarthria, dysphagia, facial and temporal atrophy, poor cough, often dystonic posturing and ophthalmoparesis, and occasionally blepharospasm. The disease progresses relentlessly; death from pulmonary complications and cachexia occurs from six to 29 years after onset [Sudarsky et al 1992, Sequeiros & Coutinho 1993]. In a recent Brazilian study, the mean age of onset was 36 years with a 21-year mean survival after onset [Kieling et al 2007].Subtypes of SCA3. Occasionally, family members with similar allele size may exhibit other clinical features such as a dystonic-rigid syndrome, a parkinsonian syndrome, or a combined syndrome of dystonia and peripheral neuropathy. Individuals with later adult onset often have a disorder that combines ataxia, generalized areflexia, and muscle wasting. Based on this phenotypic variability, Portuguese researchers classified SCA3 into several types, as reviewed recently [Riess et al 2008]. In some individuals one type can evolve into another during the course of disease [Fowler 1984]. Striving to place affected individuals into a specific SCA3 subtype has little clinical value, partly because the overlap across subtypes is considerable. The existence of subtypes, however, does illustrate the extreme clinical heterogeneity of SCA3.Type I disease (13% of individuals) is characterized by onset at a young age and prominent spasticity, rigidity, and bradykinesia, often with little ataxia [Lu et al 2004]. Type I disease is associated with longer disease-causing repeat alleles.Type II disease, the most common (57%), is characterized by ataxia and upper motor neuron signs. Spastic paraplegia can be part of the phenotype [Landau et al 2000]. Type II disease is associated with a wide range of disease-causing repeat alleles, the majority of which are in the middle of the disease range.Type III disease (30%) manifests at a later age with ataxia and peripheral polyneuropathy. Type III disease is associated with shorter disease-causing repeat alleles.Type IV disease is characterized by dopa-responsive parkinsonism; this type is not associated with any particular size disease-causing repeat alleles.Type V disease resembling hereditary spastic paraplegia has been suggested by some observers [Wang et al 2009]; however, this designation has not been commonly accepted.MRI. Brain imaging studies typically reveal pontocerebellar atrophy [Burk et al 1996]. The most commonly observed abnormality is enlargement of the fourth ventricle [Onodera et al 1998], which reflects atrophy of the cerebellum and brain stem. The degree of brain atrophy detectable by MRI varies greatly, consistent with the wide clinical variability observed. In a large European natural history study, clinical dysfunction in SCA3 correlated with the degree of total brain stem atrophy [Schulz et al 2010]. Abnormal linear high intensity of the globus pallidus interna on T₂ and FLAIR images has also been observed [Yamada et al 2005].NCV. Nerve conduction velocity studies often reveal evidence for involvement of the sensory nerves as well as the motor neurons [Lin & Soong 2002, França et al 2009]. Neuropathologic studies typically reveal neuronal loss in the pons, substantia nigra, thalamus, anterior horn cells and Clarke's column in the spinal cord, vestibular nucleus, many cranial motor nuclei, and other brain stem nuclei [Rüb et al 2002, Rüb et al 2004a, Rüb et al 2004b, Rüb et al 2006]. The cerebellum typically shows atrophy, but in some individuals Purkinje cells and inferior olivary neurons are relatively spared [Sequeiros & Coutinho 1993]. Recent neuropathologic studies have established that degeneration is rather widespread in SCA3, not confined to the cerebellum, brain stem, and basal ganglia [Rüb et al 2008]. In general, however, the cerebral cortex is spared in disease.

Genotype-Phenotype CorrelationsProbands. As with other CAG trinucleotide repeat expansion disorders, an inverse relationship exists between the age of onset and the number of CAG repeats in the abnormal allele, with the correlation coefficient ranging from -0.67 to -0.92 in a series of genotype-phenotype correlation studies performed in the 1990s. However, more recent analyses by European investigators found that only about 46%-48% of the variability in age of onset of SCA3 is accounted for by CAG repeat length, indicating that other genetic or non-genetic factors also contribute [van de Warrenburg et al 2005, Globas et al 2008]. A loose correlation also exists between the repeat number and the clinical phenotype. Individuals classified as having type I disease (dystonic-rigid form) tend to have larger repeat sizes than individuals with type II disease (ataxia with pyramidal signs) or type III disease (peripheral amyotrophy). In general, individuals with type III disease have onset later in adulthood, more prominent polyneuropathy, and 73 or fewer CAG repeats. In the study by Sasaki et al [1995], individuals with type I disease had a mean CAG repeat length of 80, those with type II disease had a mean CAG repeat length of 76, and those with type III disease had a mean CAG repeat length of 73.Some, but perhaps not all, observed anticipation in age of onset can be accounted for by the intergenerational expansion of the repeat length. Some of the largest expansions, reported by Zhou et al [1997] from China in two children with onset at age 11 and age five years, had CAG repeats of 83 and 86, respectively. The severity of disease manifestations, although related to the age of disease onset, varies among families. An example is a Yemeni family in which two groups of affected individuals were distinguished by the age of onset [Lerer et al 1996]: an obligate heterozygote died at age 70 years having had no symptoms, and another person with 68 repeats was asymptomatic at age 66 years. More severe disease has been reported in homozygous individuals in a few other families [Lerer et al 1996, Carvalho et al 2008]. However, many of the homozygotes in the Yemeni family are no more severely affected than heterozygotes in other families.

Differential DiagnosisProgressive ataxia, often associated with evidence of upper motor neuron dysfunction including brisk tendon reflexes and extensor plantar responses, can be seen in individuals with spinocerebellar ataxia type 3 (SCA3), also known as Machado-Joseph Disease (MJD), as well as in many other dominantly inherited ataxias (see Ataxia Overview).Findings suggestive of SCA3 include occurrence of variant clinical features in members of the same family, such as an akinetic-rigid syndrome (often responsive to dopaminergic agonists) or cerebellar ataxia associated with significant peripheral amyotrophy and generalized areflexia. The presence of dystonia and parkinsonian features, including a beneficial response to levodopa or dopamine agonists, can cause diagnostic confusion with dopa-responsive dystonia and Parkinson disease [Schols et al 2000]. In SCA3, however, most individuals manifesting with parkinsonian features also have some evidence of cerebellar involvement.Note to clinicians: For a patient-specific ‘simultaneous consult’ related to this disorder, go to , an interactive diagnostic decision support software tool that provides differential diagnoses based on patient findings (registration or institutional access required).

ManagementEvaluations Following Initial DiagnosisTo establish the extent of disease in an individual diagnosed with spinocerebellar ataxia type 3 (SCA3), also known as Machado-Joseph disease (MJD), the following evaluations are recommended if specific symptoms are present:MRI of the brain and spinal cord if there are cognitive problems or unusually severe ataxia, motor, or sensory findings Speech and swallowing assessment if dysarthria and/or dysphagia are present Gait assessment EMG/NCV if peripheral symptoms are present to assess the degree of involvement of the peripheral nervous system Treatment of ManifestationsManagement of individuals with SCA3 remains supportive as no medication has been proven to slow the course of disease [D’Abreu et al 2010].However, some symptoms may respond — in some cases dramatically — to certain drugs:Particularly the extrapyramidal syndromes resembling parkinsonism may benefit from levodopa or dopamine agonist [Subramony et al 1993, Nandagopal & Moorthy 2004]. Symptoms of restless legs syndrome may respond to these agents. Other manifestations, including spasticity, drooling, and sleep problems, also respond variably to appropriate agents such as lioresal, atropine-like drugs, and hypnotic agents. Botulinum toxin has been used for dystonia and spasticity [Freeman & Wszolek 2005]. Daytime fatigue, a common problem in individuals with SCA3, may respond to psychostimulants used in narcolepsy such as modafinil. Depression is common in individuals with SCA3 and should be treated with antidepressants [Cecchin et al 2007]. A trial of occupational therapy in SCA3 showed that depression scores improved as a consequence of therapy, underscoring the fact that non-pharmacologic measures may also improve affective disorder in SCA3 [Silva et al 2010].Of note, relatively few clinical trials of medications have been performed in SCA3, and none has yet been confirmed to shown a definite benefit:A small study of six individuals with SCA3 suggested that lamotrigine may improve balance; however, benefit was not confirmed during the withdrawal phase of the trial [Liu et al 2005]. While an earlier study suggested that tremethoprim-sulfamethoxazole may be beneficial in treating SCA3 [Sakai et al 1995], a larger study of 22 persons failed to show any benefit [Schulte et al 2001], leading the authors to conclude that long-term therapy with this drug combination is not recommended. A study of fluoxetine failed to show benefit for motor symptoms [Monte et al 2003]. A study of the 5-HT1A agonist, tandospirone, in ten persons suggested improvement in depressive symptoms, ataxia, insomnia, and leg pain in a subset of individuals [Takei et al 2004]. A subsequent open-label four-week symptomatic study by the same investigators tested tandospirone in a variety of degenerative ataxias, including 14 persons with SCA3. Four of 14 showed improved scores in an ataxia rating scale [Takei et al 2010]. A larger, double blind placebo-controlled study is needed to confirm such benefits.Non-pharmacologic therapy is important in SCA3:Although neither exercise nor physical therapy slows the progression of incoordination or muscle weakness, affected individuals should maintain activity. Canes and walkers help prevent falling, and motorized scooters later in disease can help maintain independence. Speech therapy and communication devices such as writing pads and computer-based devices may benefit those with dysarthria. When dysphagia becomes troublesome, video esophagrams can identify the consistency of food least likely to trigger aspiration. Modification of the home with such conveniences as grab bars, raised toilet seats, and ramps to accommodate motorized chairs may be necessary. Weighted eating utensils and dressing hooks help maintain a sense of independence. Prevention of Secondary ComplicationsVitamin supplements are recommended, particularly if caloric intake is reduced.Weight control is important because obesity can exacerbate difficulties with ambulation and mobility.General anesthesia may be problematic; experience with local anesthesia has been reported [Teo et al 2004].SurveillanceSpeech, swallowing, and gait function should be monitored annually or biannually to assess need for assistive devices.Evaluation of Relatives at RiskSee Genetic Counseling for issues related to testing of at-risk relatives for genetic counseling purposes.Therapies Under InvestigationSearch ClinicalTrials.gov for access to information on clinical studies for a wide range of diseases and conditions. Note: There may not be clinical trials for this disorder.OtherTremor-controlling drugs do not work well for cerebellar tremors.No dietary factor has been shown to curtail symptoms.

Molecular GeneticsInformation in the Molecular Genetics and OMIM tables may differ from that elsewhere in the GeneReview: tables may contain more recent information. —ED.Table A. Spinocerebellar Ataxia Type 3: Genes and DatabasesView in own windowGene SymbolChromosomal LocusProtein NameLocus SpecificHGMDATXN314q32​.12Ataxin-3ATXN3 homepage - Mendelian genesATXN3Data are compiled from the following standard references: gene symbol from HGNC; chromosomal locus, locus name, critical region, complementation group from OMIM; protein name from UniProt. For a description of databases (Locus Specific, HGMD) to which links are provided, click here.Table B. OMIM Entries for Spinocerebellar Ataxia Type 3 (View All in OMIM) View in own window 109150MACHADO-JOSEPH DISEASE; MJD 607047ATAXIN 3; ATXN3Normal allelic variants. ATXN3 consists of 11 exons within a 1776-base pair coding region with one long open reading frame [Kawaguchi et al 1994, Ichikawa et al 2001]. A normal variation in the CAG trinucleotide repeat encoding a polyglutamine repeat occurs within exon 10. Variations in the (CAG)_n sequence exist. It can be an imperfect repeat in many alleles examined, where the third, fourth, and sixth CAG unit is replaced by CAA, AAG, and CAA, respectively. These variant triplets were commonly found both in normal and abnormal alleles. The CAG repeat is highly variable in normal individuals with the (CAG)_n in different alleles varying from 12 to 43 [Cancel et al 1995, Maciel et al 1995, Matilla et al 1995, Ranum et al 1995, Sasaki et al 1995, Takiyama et al 1995, Limprasert et al 1996, Matsumura et al 1996a. In many studies, the distribution of CAG repeat numbers in normal alleles has shown a bimodal or trimodal pattern with peaks around 14, 22-24, and 27. Rubinsztein et al [1995] looked at 748 normal alleles from individuals of eight different ethnic backgrounds who had spinocerebellar ataxia type 3 (SCA3, MJD) and found a similar bimodal distribution of normal CAG repeat numbers with peaks at 14 and 21-23. The proportion of heterozygotes varied among populations, with higher figures among Melanesians (98%), Polynesians (92%), and African blacks (88%), and the fewest among East Anglicans (58%). Limprasert et al [1996] found 14 and 23 (CAG)_n repeat lengths to be the most common across populations. Furthermore, in analyzing the (CAG)_n tracts in different species, including humans, it was observed that usually the CAG repeat was flanked with a guanine, but when the alleles had 20 or 21 CAG repeats, the guanine was replaced with cytosine. In addition, cytosine occurred in 54.5% of normal alleles with CAG repeat numbers between 27 and 40, with a frequency distribution of 23%-100% among different ethnic populations. All expanded alleles also contained cytosine at the first nucleic acid residue following the (CAG)_n tracts. Cytosine at this point may play a role in determining the instability of polyglutamine tracts [Matsumura et al 1996b]. Overall, 93.5% of normal chromosomes have fewer than 31 CAG repeats. Pathologic allelic variants. SCA3 (MJD) is caused by abnormally large number of CAG repeats [Kawaguchi et al 1994, Cancel et al 1995, Maciel et al 1995, Matilla et al 1995, Ranum et al 1995, Takiyama et al 1995, Schols et al 1996, Silveira et al 1996, Takiyama et al 1997]. Alleles with an abnormal number of CAG repeats may display both somatic and gametic instability of the repeat [Hashida et al 1997]. Typically, spermatozoa contain a larger repeat length than leukocytes in the same individuals [Watanabe et al 1996]. In the CNS, cerebellar tissues often tend to have slightly smaller repeat lengths than other regions of the brain. Haplotype analysis has revealed that individuals with SCA3 from different populations often shared the same haplotype, suggesting presence of a founder effect [Takiyama et al 1995]. However, in the restricted populations of the Azores, two distinct haplotypes have been found, a fact that could overrule the one founder mutation theory [Gaspar et al 1996]. Mittal et al [2005] found evidence for the single Portuguese founder allele in India. Table 2. Selected ATXN3 Allelic Variants View in own windowClass of Variant AlleleDNA Nucleotide Change
(Alias ¹) Protein Amino Acid Change Reference SequencesNormalc.886_888CAG(<44)
(<44 CAG repeats) p.Gln296(<44)NM_004993​.5
NP_004984​.2Reduced penetrancec.886_888CAG(45_51)
(45 to 51 CAG repeats)p.Gln296(45_51)Pathologicc.886_888CAG(52_86)
(52 to 86 CAG repeats)p.Gln296(52_86)See Quick Reference for an explanation of nomenclature. GeneReviews follows the standard naming conventions of the Human Genome Variation Society (www​.hgvs.org).1. Variant designation that does not conform to current naming conventionsNormal gene product. Ataxin-3, the protein encoded by ATXN3, is a de-ubiquitinating enzyme with a predicted molecular weight of approximately 42 kd. It is widely expressed in the brain and throughout the body, existing both in the cytoplasm and nucleus of various cell types. In neurons, ataxin-3 is predominantly a cytoplasmic protein [Paulson et al 1997a], but the protein readily shuttles in and out of the nucleus. Ataxin-3 is a ubiquitin-specific protease that binds and cleaves ubiquitin chains [Burnett et al 2003, Donaldson et al 2003, Doss-Pepe et al 2003, Chai et al 2004, Berke et al 2005]. A highly conserved amino-terminal "josephin" domain contains the catalytic triad of amino acids found in cysteine proteases [Nicastro et al 2005]. The carboxy-terminal region of the protein contains several ubiquitin-interacting motifs (UIMs) through which the protein binds tightly to polyubiquitin chains. The polyglutamine tract resides between UIMs 2 and 3, but how the polyglutamine tract affects the enzymatic function of the protein (if at all) is unknown. Studies suggest that ataxin-3 is intrinsically prone to self-associate. The stability and aggregating properties of ataxin-3 are related both to its globular josephin domain [Chow et al 2004a, Chow et al 2004b, Masino et al 2004, Ellisdon et al 2007] and to the polyglutamine domain that is expanded in disease.Normal human ataxin-3 suppresses polyglutamine neurodegeneration in Drosophila through a mechanism that requires its ubiquitin-associated functions [Warrick et al 2005]. Ataxin-3 also regulates aggresome formation, the degradation of proteins sent from the endoplasmic reticulum and the activity of some ubiquitin ligases [Burnett & Pittman 2005, Durcan et al 2011]. Taken together with the enzymatic properties of ataxin-3, these facts suggest that ataxin-3 normally participates in various ubiquitin-dependent protein quality control pathways in the cell.Abnormal gene product. Most evidence favors a toxic protein mechanism of disease in which expansion of the polyglutamine tract in ataxin-3 makes the protein highly susceptible to misfolding and aggregation [Williams & Paulson 2008]. This process has been successfully modeled in in vitro studies employing recombinant protein and in various cellular models and transgenic animal models [Paulson et al 1997b, Warrick et al 1998, Evert et al 1999, Cemal et al 2002, Evert et al 2006, Bichelmeier et al 2007, Chen et al 2008, Alves et al 2008, Williams et al 2009, Boy et al 2009, Mueller et al 2009, Reina et al 2010]. In humans with SCA3, normal and expanded (i.e., pathogenic) ataxin-3 are widely expressed both in the brain and in other unaffected organ systems [Paulson et al 1997b]. The subcellular distribution of ataxin-3 differs in disease brain versus normal brain: normally a predominantly cytoplasmic protein in neurons, ataxin-3 becomes concentrated in the nucleus of neurons during disease. Moreover, in many brain regions, ataxin-3 forms intranuclear inclusions [Paulson et al 1997b]. These neuronal inclusions, which are also found in other polyglutamine disorders, are particularly abundant in pontine neurons but are also seen in several other brain regions. Inclusions are heavily ubiquitinated and contain heat shock molecular chaperones and proteasomal subunits, suggesting that they are repositories for aberrantly folded, aggregated protein [Schmidt et al 2002]. Axonal inclusions have also been noted [Seidel et al 2010]. Current evidence, including recent neuropathologic studies [Rüb et al 2006], suggest that inclusions are not directly pathogenic structures and may instead be the byproduct of neuronal efforts to wall off abnormal proteins [Williams & Paulson 2008]. At the very least, inclusions are a biomarker of abnormal protein accumulation and impaired protein clearance. The disease protein also forms abnormal deposits outside the cell nucleus, including in neuronal projections; whether these are direly pathogenic is also unknown. In several other disorders caused by polyglutamine expansions, cleavage of the disease protein to produce a "toxic fragment" has been postulated to contribute to pathogenesis. In SCA3, some findings support this view. For example, transgenic mice or flies expressing an ataxin-3 polyglutamine fragment show marked degeneration, much more so than flies and mice expressing the full-length mutant protein [Ikeda et al 1996, Warrick et al 1998, Warrick et al 2005]. Moreover, a putative cleavage fragment has been shown to accumulate in one mouse model of disease [Goti et al 2004], and a specific cleavage fragment also accumulates in cells undergoing apoptosis [Berke et al 2004]. Recent studies in a fly model of disease further suggest that proteolysis contributes to pathogenesis [Jung et al 2009]. Although expanded polyglutamine is favored as the primary toxic element in disease, it is possible that CAG expansion promotes ribosomal frameshifting during translation of the disease protein [Toulouse et al 2005], potentially leading to alternative protein sequences containing polyalanine tracts instead of polyglutamine tracts. Studies in Drosophila further suggest that, independent of proteotoxic effects, the expanded CAG repeat also may exert toxicity at the RNA level [Li et al 2008].