COMMENTARY: Chaperoning against neuronal vulnerability (Commentary on Zijlstra et al.)


One of the major questions in neurobiology is why particular neurons are vulnerable to neurodegenerative disease. Polyglutamine (polyQ) repeat expansions lead to a range of neurodegenerative diseases, including Huntington’s disease and the spinocerebellar ataxias. The investigation of these polyQ diseases has offered clear insights into the pathogenesis of neurodegeneration. For example, there is a strong relationship between the size of polyQ expansion and the propensity of the protein to misfold and aggregate. Larger, more aggregation-prone polyQ leads to an earlier onset of disease, suggesting that protein misfolding and aggregation might be critical in disease pathogenesis. In this issue of EJN, Zijlstra et al. illuminate this issue further by suggesting that the level of some heat shock proteins (HSPs) might also predict disease onset.

Maintaining protein homeostasis (proteostasis) requires the co-ordination of protein synthesis, folding, trafficking, disaggregation and degradation. Most HSPs function as molecular chaperones to facilitate protein folding, or integrate protein quality control with degradation, such that they are central players in the proteostasis network and the cellular response to proteostasis imbalance (Voisine et al., 2010). Several classes of molecular chaperone can modulate polyQ misfolding, aggregation and toxicity in cell and animal models (Williams & Paulson, 2008). The reduction of some chaperones enhances polyQ toxicity, whereas overexpression can protect against polyQ. Therefore, the steady-state level of chaperone proteins and the cell’s ability to respond to challenges of proteostasis might affect the toxicity of polyQ proteins in patients.

In this issue of EJN, Zijlstra et al. tested this hypothesis by investigating the expression of chaperones, which have been suggested to affect polyQ aggregation and/or toxicity in fibroblasts from patients with spinocerebellar ataxia 3, and correlated this with the age of onset. They discovered that Hsp27 (HSPB1) protein was upregulated in cells of patients with spinocerebellar ataxia 3, and suggest that this is a response to the presence of the mutant protein. Importantly, using an antibody against DNAJB1, they observed that patients with earlier disease onset (once polyQ length was accounted for) had lower levels of DNAJB1 immunoreactivity. When they checked the specificity of the commercial DNAJB1 antisera against a panel of DnaJ proteins, it was revealed to cross-react with other DnaJ proteins, albeit at lower affinity. Furthermore, DNAJB1 was more effective against ataxin 3 polyQ aggregation than the other immunoreactive DnaJ proteins of a similar molecular weight, suggesting that it is most likely to be the lower levels of DNAJB1 that are associated with accelerated spinocerebellar ataxia 3 disease pathogenesis.

The DnaJ (Hsp40) proteins are cochaperones that regulate the activity of Hsp70 proteins. They are the largest and most diverse family of chaperones, with up to 50 family members in humans. Some have neuronal enriched expression and are potent suppressors of polyQ aggregation (e.g. Westhoff et al., 2005; Hageman et al., 2010). Given their diversity and differences in expression, different DnaJ proteins could be effective in distinct brain regions against varied aggregation-prone client proteins, such as different polyQ proteins. For example, DNAJB1 was more effective against ataxin 3 polyQ than DNAJB4 and DNAJB5, whereas they had similar activity against Htt exon 1 polyQ (Hageman et al., 2010; Zijlstra et al., 2010). This suggests that it might be important to analyse specific chaperone/client interactions rather than generic interactions.

The use of fibroblasts to study gene expression in neurodegeneration may seem a little unorthodox to many neuroscientists, but there are several advantages. These patient cells are not affected by disease-related change associated with neuronal dysfunction, death or agonal state, which can confound many studies of these stress-related genes from post-mortem samples. Therefore, they may be useful to reveal genetic and epigenetic factors that regulate gene expression. However, they cannot reflect neuron-specific gene expression and in the future we will need to investigate the factors that regulate chaperone gene expression in neurons, possibly exploiting induced pluripotent stem cell (iPS) technology in conjunction with transgenic animals. Interestingly, mRNA expression did not correlate well with protein levels in these fibroblasts, so there might also be post-transcriptional control of chaperone levels and activity as part of the proteostasis network. Although the beneficial effects of chaperone expression in fly and worm models of polyQ disease are undoubted, their effects in rodent models have been less convincing. The investigation of chaperones in specific neuronal populations and diseases in mammals may help to identify the best predictors of neuronal vulnerability to proteostasis imbalance and the most effective therapeutic targets.