Expanding Our Understanding of Polyglutamine Diseases through Mouse Models

becomes significant only after 6 months of age, long The polyglutamine neurodegenerative diseases—a group after the phenotype has appeared. The axons of the that so far includes spinobulbar muscular atrophy murine Purkinje cells are not dilated like the human (SBMA), Huntington’s disease (HD), dentatorubropalliSCA1-affected cells, but this is the only pathologic difdoluysian atrophy (DRPLA) and the spinocerebellar ference between the two. The B05 mice thus provide a ataxias (SCA1, 2, 3, 6, and 7)—result from the expansion reliable animal model for SCA1—but leave unanswered of an unstable CAG trinucleotide repeat coding for polythe question whether it is the full-length mutant protein glutamine tracts in the respective proteins. The CAG or merely the expanded glutamine repeat that induces repeat tracts are polymorphic and vary in their instability: polyglutamine diseases. the SCA7 gene, for example, may expand by hundreds of Polyglutamine Tract or Full-Length Protein? repeats in one intergenerational transmission, whereas The first attempts to answer this question were to come the SCA6 gene is relatively stable and the difference in from SCA3 transgenic mice. SCA3, also known as Malength between the wild-type and mutant CAG tract is chado-Joseph Disease (MJD), is similar to SCA1 except only three repeats. But in all these diseases, triplet rethat dentate neurons, rather than Purkinje cells, are the peats exceeding their normal range cause progressive primary sites of cerebellar pathology. Basal ganglia neuronal dysfunction and death within 10–20 years after involvement is also prominent. Ikeda and colleagues the onset of symptoms, with longer repeat tracts caus(1996) used Pcp2/L7 to generate transgenic mice exing earlier age of onset and more severe disease. Curipressing either full-length or truncated versions of ously, only a specific subset of neurons is vulnerable in ataxin-3. Mice expressing full-length ataxin-3 with 79 each of these diseases, despite the ubiquitous expresglutamines, designated MJD79, developed no ataxia or sion of the relevant disease proteins throughout the brain pathological changes; neither did Q35C mice, which exand other tissues. The normal functions of these proteins pressed a truncated form of ataxin-3 with a CAG tract remain a mystery, except for the androgen receptor of 35 repeats. In contrast, both Q79C mice (carrying (SBMA) and the a1A voltage-dependent calcium channel truncated ataxin-3 with 42 amino acids C terminal to a (SCA6). The importance of mouse models for studying CAG tract of 79 repeats) and Q79 mice (bearing a peptide polyglutamine pathogenesis will become apparent after containing 79 glutamines) became ataxic by 4 weeks of a brief look at the issues that have become clearer, and age. By 8 weeks, Q79C mice showed massive degenerain some cases more complex, through their use. tion of all three layers of the cerebellum, which was The First Mouse Model for a Polyglutamine Disease reduced to about one-eighth of its normal volume. Cerebellar atrophy with severe Purkinje cell degeneraIt is noteworthy that the expanded polyglutamine tract tion is responsible for the SCA1 ataxic phenotype is sufficient to cause neuronal death, and that its toxicity (Zoghbi and Ballabio, 1995). Wanting to study the Purcauses massive cell loss as opposed to the more prokinje cell pathology at the root of this disease, Burright gressive dysfunction seen, for example, with ataxin-1. et al. (1995) expressed full-length human SCA1 cDNAs Ikeda and colleagues (1996) proposed that cell-specific with different numbers of CAG repeats in Purkinje cells proteolytic cleavage of the mutant protein frees the toxic using a Purkinje cell–specific promoter from the Pcp2/ polyglutamine tract, which then induces cell death. L7 gene. The resulting mice highly express either a wildThere is no evidence yet, however, that ataxin-3 cleavtype SCA1 allele with 30 repeats (30Q) or an expanded age occurs in affected brain regions of MJD/SCA3 paallele with 82 repeats (82Q). The 30Q mice are indistintients. Moreover, it is impossible to conclude that fullguishable from wild-type littermates, but adult 82Q mice length ataxin-3 is not toxic without data on transgene (known as the B05 line) develop severe ataxia and proexpression levels from this series of mice. gressive Purkinje cell pathology (Clark et al., 1997). During the same period, Mangiarini and colleagues The first histologic abnormalities, detectable at P25, (1996) generated mice using a 1.9 kb human genomic are cytoplasmic vacuoles. Loss of proximal dendritic fragment containing huntingtin (HD) 59 flanking searborization and dendritic spines becomes apparent at quences and exon 1 with an unstable expanded CAG 5 weeks when the mice begin to show mild difficulty on tract of z130 repeats. The R6/2 line (with 144 repeats) the rotating rod; by the time 82Q mice are overtly ataxic ubiquitously expresses the first 69 amino acids of hun(12–15 weeks), the dendritic arborization is mostly lost, tingtin with the elongated CAG tract at lower-thanthe molecular layer is atrophied, and some heterotopic endogenous levels. Expecting to study CAG repeat in-