Christopher J. Cummings

Chaperone suppression of ataxin-1 aggregation and altered subcellular proteasome localization imply protein misfolding in SCA1

Spinocerebellar ataxia type 1 (SCA1) is an autosomal dominant neurodegenerative disorder caused by expansion of a polyglutamine tract in ataxin-1. In affected neurons of SCA1 patients and transgenic mice, mutant ataxin-1 accumulates in a single, ubiquitin-positive nuclear inclusion. In this study, we show that these inclusions stain positively for the 20S proteasome and the molecular chaperone HDJ-2/HSDJ. Similarly, HeLa cells transfected with mutant ataxin-1 develop nuclear aggregates which colocalize with the 20S proteasome and endogenous HDJ-2/HSDJ. Overexpression of wild-type HDJ-2/HSDJ in HeLa cells decreases the frequency of ataxin-1 aggregation. These data suggest that protein misfolding is responsible for the nuclear aggregates seen in SCA1, and that overexpression of a DnaJ chaperone promotes the recognition of a misfolded polyglutamine repeat protein, allowing its refolding and/or ubiquitin-dependent degradation.

Ataxin-1 with an expanded glutamine tract alters nuclear matrix-associated structures

Spinocerebellar ataxia type 1 (SCA1) is one of several neurodegenerative disorders caused by an expansion of a polyglutamine tract. It is characterized by ataxia, progressive motor deterioration, and loss of cerebellar Purkinje cells. To understand the pathogenesis of SCA1, we examined the subcellular localization of wild-type human ataxin-1 (the protein encoded by the SCA1 gene) and mutant ataxin-1 in the Purkinje cells of transgenic mice. We found that ataxin-1 localizes to the nuclei of cerebellar Purkinje cells. Normal ataxin-1 localizes to several nuclear structures ∼0.5 µm across, whereas the expanded ataxin-1 localizes to a single ∼2-µm structure, before the onset of ataxia. Mutant ataxin-1 localizes to a single nuclear structure in affected neurons of SCA1 patients. Similarly, COS-1 cells transfected with wild-type or mutant ataxin-1 show a similar pattern of nuclear localization; with expanded ataxin-1 occurring in larger structures that are fewer in number than those of normal ataxin-1. Colocalization studies show that mutant ataxin-1 causes a specific redistribution of the nuclear matrix-associated domain containing promyelocytic leukaemia protein. Nuclear matrix preparations demonstrate that ataxin-1 associates with the nuclear matrix in Purkinje and COS cells. We therefore propose that a critical aspect of SCA1 pathogenesis involves the disruption of a nuclear matrix-associated domain.

Mutation of the E6-AP Ubiquitin Ligase Reduces Nuclear Inclusion Frequency While Accelerating Polyglutamine-Induced Pathology in SCA1 Mice

Mutant ataxin-1, the expanded polyglutamine protein causing spinocerebellar ataxia type 1 (SCA1), aggregates in ubiquitin-positive nuclear inclusions (NI) that alter proteasome distribution in affected SCA1 patient neurons. Here, we observed that ataxin-1 is degraded by the ubiquitin-proteasome pathway. While ataxin-1 [2Q] and mutant ataxin-1 [92Q] are polyubiquitinated equally well in vitro, the mutant form is three times more resistant to degradation. Inhibiting proteasomal degradation promotes ataxin-1 aggregation in transfected cells. And in mice, Purkinje cells that express mutant ataxin-1 but not a ubiquitin-protein ligase have significantly fewer NIs. Nonetheless, the Purkinje cell pathology is markedly worse than that of SCA1 mice. Taken together, NIs are not necessary to induce neurodegeneration, but impaired proteasomal degradation of mutant ataxin-1 may contribute to SCA1 pathogenesis.

The cerebellar leucine-rich acidic nuclear protein interacts with ataxin-1

Spinocerebellar ataxia type 1 (SCA1) is an autosomal dominant neurodegenerative disorder characterized by ataxia, progressive motor deterioration, and loss of cerebellar Purkinje cells. SCA1 belongs to a growing group of neurodegenerative disorders caused by expansion of CAG repeats, which encode glutamine. Although the proteins containing these repeats are widely expressed, the neurodegeneration in SCA1 and other polyglutamine diseases selectively involves a few neuronal subtypes. The mechanism(s) underlying this neuronal specificity is unknown. Here we show that the cerebellar leucine-rich acidic nuclear protein (LANP) interacts with ataxin-1, the SCA1 gene product. LANP is expressed predominantly in Purkinje cells, the primary site of pathology in SCA1. The interaction between LANP and ataxin-1 is significantly stronger when the number of glutamines is increased. Immunofluorescence studies demonstrate that both LANP and ataxin-1 colocalize in nuclear matrix-associated subnuclear structures. The features of the interaction between ataxin-1 and LANP, their spatial and temporal patterns of expression, and the colocalization studies indicate that cerebellar LANP is involved in the pathogenesis of SCA1.

Expanding Our Understanding of Polyglutamine Diseases through Mouse Models

The above mouse models strongly support the notion that polyglutamine pathogenesis is caused by some toxic function gained by the mutant protein as a result of the expanded CAG repeat. Selective neuronal vulnerability may be determined by the protein context of the expanded polyglutamine tract and the level of protein in those cells. Mutant proteins may be abnormally processed or interact abnormally with other proteins, which presumably contribute to selective neuronal dysfunction and degeneration (Figure 1Figure 1).Figure 1Model for the Pathogenesis of Polyglutamine DiseasesView Large Image | View Hi-Res Image | Download PowerPoint SlideIs the pathogenesis of all polyglutamine disorders mediated by the same mechanism(s)? The answer is most likely “yes and no.” Protein misfolding and turnover seem to be shared features, since glutamine tract expansion leads to the accumulation of mutant protein in all the diseases evaluated. The variation in sites of accumulation of the mutant protein, however—in SCA2 and SCA6, the mutant proteins accumulate in the cytoplasm, not the nucleus (6xHuynh, D.P, Del Bigio, M.R, Ho, D.H, and Pulst, S.M. Ann. Neurol. 1999; 45: 232–241Crossref | PubMed | Scopus (114)See all References, 8xIshikawa, K, Fujigasaki, H, Saegusa, H, Ohwada, K, Fujita, T, Iwamoto, H, Komatsuzaki, Y, Toru, S, Toriyama, H, Watanabe, M et al. Hum. Mol. Genet. 1999; 8: 1185–1193Crossref | PubMed | Scopus (147)See all References)—implies that the downstream pathogenic mechanisms may differ. Polyglutamine expansion could cause the mutant proteins to adopt altered conformations, leading to their ubiquitination, aggregation, and resistance to proteasomal degradation. Over the course of the disease, the ubiquitin–proteasome system could become encumbered by the aggregate-prone proteins, which in turn could alter the turnover of other critically short-lived proteins. Although neuronal dysfunction and cytoplasmic changes could occur very early in the pathogenesis, later-stage disruptions of nuclear structures and/or functions might be common factors in several polyglutamine diseases.Even a unifying model for polyglutamine-induced neurodegeneration, however, leaves a number of questions unanswered including the following key issues. What cell-specific determinants underlie the selectivity of neuronal degeneration: cell-specific interactors or cell-specific alterations of gene expression? Are the functions gained by the mutant proteins novel or intrinsic? What are the earliest physiological changes in affected neurons? Is there therapeutic potential in inhibiting cleavage or nuclear transport of the mutant proteins?Present and future generation animal models—mouse, Drosophila, and C. elegans alike—will be invaluable tools in addressing many of these questions. For example, the significance of interacting proteins in selective neuronal vulnerability can now be studied by mating the various models with animals that either lack or overexpress the interactors in the vulnerable neurons. The emerging principle that truncation of the parent protein is an important step in the pathogenesis of several of these diseases can now be tested by investigating the posttranslational processing of these proteins in vivo. Ona et al. 1999xOna, V.O, Li, M, Vonsattel, J.P, Andrews, L.J, Khan, S.Q, Chung, W.M, Frey, A.S, Menon, A.S, Li, X.J, Stieg, P.E et al. Nature. 1999; 399: 263–267Crossref | PubMed | Scopus (496)See all ReferencesOna et al. 1999, for example, recently showed that caspase-1 is activated in the R6/2 mice, and inhibiting its activity reduces endogenous huntingtin cleavage. Furthermore, this causes several of the phenotypes described above to progress more slowly. It would be interesting to see if inhibition of caspase-1 would also modify the phenotype of mice expressing full-length huntingtin. Lessons learned from in vitro studies—e.g., ataxin-1 aggregation in transfected HeLa cells can be suppressed by overexpressing a molecular chaperone—can now be carried into mouse models to determine whether these modifications occur in vivo. Likewise, the putative involvement of the ubiquitin–proteasome pathway might be investigated by mating the polyglutamine mouse models with mice that have deficiencies in this pathway or, conversely, with transgenic mice overexpressing factors within the pathway. Most importantly, mouse models allow the study of early phenotypic changes for which patient material is seldom available. Gene chips and other array technologies should prove invaluable for finding transcriptional changes early in the disease process. Characterizing changes that occur prior to any pathological events will greatly expand our knowledge of the molecular mechanisms that cause neurodegeneration in polyglutamine diseases.*To whom correspondence should be addressed (e-mail: hzoghbi@bcm.tmc.edu).

Correction: The cerebellar leucine-rich acidic nuclear protein interacts with ataxin-1

This corrects the article DOI: 10.1038/40159

Erratum: The cerebellar leucine-rich acidic nuclear protein interacts with ataxin-1 (Nature (1997) 389 (974-978))

This corrects the article DOI: 10.1038/40159

Erratum: Ataxin-1 with an expanded glutamine tract alters nuclear matrix-associated structures (Nature (1997) 389 (971-974))

This corrects the article DOI: 10.1038/40153