Most of the polyQ disease proteins are widely expressed within and outside of the brain. However, expansion of the polyQ tract results in an essentially neuronal-specific phenotype in patients. Moreover, each of the polyQ diseases is distinguished by a unique profile of selective neurodegeneration that can be evinced radiographically or by postmortem analysis (Fig. 1). Although the pathological relevance remains controversial, the polyQ disease brain is characterized by the presence of aggregates or inclusions. These structures, the subcellular localization of which depends on the polyQ disease protein, generally have not been observed outside the central nervous system. In SBMA (Li et al. 1998) and SCA7 (Jonasson et al. 2002), however, there is evidence of nuclear inclusions in certain peripheral tissues. Nuclear aggregation in neuronal tissue is prominent in all of the polyQ diseases except SCA2 (Huynh et al. 1999) and SCA6 (Ishikawa et al. 1999, 2001). Cytoplasmic aggregates are present in some of the polyQ diseases, including SCA2 (Huynh et al. 2000), SCA6 (Ishikawa et al. 1999), HD (DiFiglia et al. 1997; Gutekunst et al. 1999), DRPLA (Hayashi et al. 1998a,b), and SMBA (Adachi et al. 2005). In HD, cy-toplasmic aggregates, which primarily localize in neuronal processes such as axons and dendrites, have been extensively characterized (DiFiglia et al. 1997; Gutekunst et al. 1999).
Mouse models of polyQ diseases have proven valuable for study of nuclear accumulation of mutant polyQ proteins in neuronal tissues. In knockin and transgenic mouse models, it is often possible to arrange these distinct labeling patterns in a histochemical time course (Michalik and Broeckhoven 2003). Diffuse nuclear staining, which increases in intensity with age and probably represents the presence of abundant microaggregates, is the initial histological event. Multiple puncta eventually become discernable within the diffuse immunoreactive signal. Ultimately, loss of the diffuse staining pattern is coincident with the emergence of, in most cases, a single neuronal intranuclear inclusion (NII) (Yvert et al. 2000; Schilling et al. 2001). The duration of each phase within this time course is directly related to the length of the polyQ tract in the transgene-encoded protein. Comparison of two different lines of SCA17 transgenic mice of the same genetic background, which express TBP with a polyQ tract of either 71 or 105 residues under a prion protein promoter, demonstrates this relationship. At 2-2.5 months of age, NII are detected prominently in the cerebellum of 105Q female mice, whereas immunoreactive cerebellar neurons in identically aged 71Q female mice are characterized by diffuse albeit intense nuclear staining with occasional puncta (M. Friedman and X.J.Li, unpublished data). In-triguingly, a conditional mouse model of HD has provided evidence that polyQ-mediated neuropathology may be reversible. In this model, aggregate formation and pathology in striatal neurons are contingent on the continued production of a mutant htt fragment (Yamamoto et al. 2000).
Although nuclear accumulation of mutant protein has received considerable attention as a neuropathological feature of the polyQ diseases, the relevance of this phenomenon to neurodegeneration is not entirely clear. Neurons containing nuclear aggregates are not necessarily prevalent in the brain regions that selectively degenerate in a given polyQ disease. Conversely, nuclear aggregates can be abundant in mildly affected or even unaffected areas of the brain. In HD, for example, neuronal death is most prominent in the caudate and putamen, but intranuclear inclusions are sparse in the striatum of patients. Aggregates abound in the lesser-affected HD cortex, however (Gutekunst et al. 1999). Moreover, within the striatum, aggregates are rarely observed in medium spiny neurons, which are selectively degraded in HD, but are prevalent in spared interneurons (Kuemmerle et al. 1999). Neu-ropathological evaluation of postmortem brains from patients with SCA17 as well as other polyQ diseases has revealed similar discrepancies (Fuji-gasaki et al. 2001; Adachi et al. 2005; Yamada et al. 2001). Interestingly, in most transgenic mouse models, despite the rapid appearance of aggregates due to overexpression of a particular polyQ disease protein or a fragment thereof, neurodegeneration is absent or not obvious (Clark et al. 1997; Abel et al. 2001; Schilling et al. 1999a,b; Mangiarini et al. 1996; Ordway et al. 1997). The short life span of mice may limit the extent of neurodegeneration, which can precede symptom manifestation in patients with polyQ disease (Albin et al. 1992) but generally becomes pronounced in the late stages of pathology.
Was this article helpful?