Surprisingly, the benefits afforded by HSP990 were transient. Rotarod performance of R6/2 HD mice improved with treatment at 8 and 10 weeks of age, but not at 14 weeks, which suggests that impairment of the HSR might occur upon HD progression (10
). This decline in behavior with HD progression was also reflected in Hsp expression levels. In young mice, HSP990-induced upregulation of Hsp70, Hsp40, and Hsp25 was comparable in WT and R6/2 mice. However, by 8 weeks of age, impairment in Hsp induction was already apparent for the R6/2 mice, and this deterioration continued with age until the mice reached end-state disease at 15 weeks (Figure and ref. 10
). In these late-stage disease R6/2 mice, treatment with HSP990 failed to induce Hsp expression. Thus, HSP990 efficacy declined sharply in R6/2 mice in an age-dependent manner. Importantly, this impairment of the HSR was not an artifact of the R6/2 model of HD. The HSR also became impaired in late-stage Hdh
Q150 knockin mice, which model late-onset HD and express full-length human huntingtin with a polyglutamine tract of 150 residues. Taken together, these data suggest it will be important to determine whether the HSR also becomes impaired in the brains of HD patients.
Further studies revealed that HSR impairment occurred at the level of transcription (10
). This impaired transcription could have several potential origins. To induce Hsp expression, HSF1 must dissociate from its repressive complex with Hsp90, translocate to the nucleus, and be hyperphosphorylated (Figure ). Coimmunoprecipitation studies indicated that HSP990 effectively dissociated Hsp90 and HSF1 even in older R6/2 mice (10
). Additionally, after HSP990 treatment, HSF1 was hyperphosphorylated and localized to the nucleus in R6/2 brain tissue. Thus, HSF1 is activated by HSP990 equally well in WT and R6/2 mice. However, the authors further noted that alterations in chromatin architecture in older R6/2 mice precluded HSF1 from engaging the promoters of heat shock genes. As the R6/2 mice aged, lower levels of HSF1 binding at various heat shock gene promoters was observed upon HSP990 treatment. Furthermore, chromatin immunoprecipitation using the Hsp70
promoter revealed lower levels of associated RNA polymerase II in R6/2 than in WT brain tissue, consistent with reduced transcription. Additionally, hypoacetylation of histone H4 was observed at various heat shock genes as disease progressed (10
), which might reduce the ability of HSF1 to bind target promoters by reducing chromatin accessibility (Figure ). However, additional experiments showed that the Hsp70
promoter region was equally accessible in both WT and R6/2 mice. Thus, accessibility per se does not appear to be the issue.
Labbadia et al. propose a model whereby as HD progresses, histone H4 becomes hypoacetylated at heat shock gene promoters, perhaps because polyglutamine aggregates sequester key histone acetyltransferases (10
). Through an undefined mechanism, hypoacetylation precludes HSF1 binding to heat shock gene promoters. Thus, the HSR is impaired (Figure ). The discovery that the HSR becomes impaired as disease progresses in HD model mice is perhaps the authors’ most significant finding. The reduced ability to launch the HSR is likely to exacerbate HD progression by rendering cells incapable of responding appropriately to environmental stressors. However, it remains unknown whether similar defects in the HSR occur in HD patients. Thus, it is possible (although unlikely, in our view) that these deficits reflect events unique to mouse models of HD that are unrelated to events in HD patients. It will be important, although challenging, to corroborate these findings in HD patient tissue or cell lines. Of note, the authors did not generate survival curves (10
), so the therapeutic value of HSP990 treatment with regard to longevity remains unknown. However, even a transient improvement of disease phenotype may bring an enhanced quality of life to HD patients, regardless of any potential increase in lifespan.