Role of molecular chaperones in protein folding diseases

点击翻译按钮获取中文摘要

The most widely studied yeast prions are the PSI+, URE3, and PIN+, which are the misfolded protein forms of Sup35, Ure2, and Rnq1, respectively. Although condensate formation of properly folded Sup35 has been established, it is not clear whether the properly folded Ure2 and Rnq1 form condensates under the same conditions as Sup35 since aside from their prion domain, the other domains are not conserved. Similar to Sup35, Ure2 and Rnq1 form condensates under starvation conditions that simultaneously deplete ATP and lower pH. The condensates formed from the three prion proteins are readily reversible upon addition of upon addition of glucose, which restores both ATP and pH to physiological levels. However, when pH is raised without replenishing ATP, only Ure2 and Rnq1 condensates readily dissolve. Furthermore, depletion of ATP at physiological pH produced condensates of Sup35, whereas the Rnq1 and Ure2 proteins remain highly mobile under these conditions and do not form condensates. Interestingly only Sup35 has a M-domain, which has been reported to be a pH sensor domain that inhibits condensate formation at higher pH. These results suggest that the sensing of pH by prion proteins and the conditions for formation and dissolution of prion condensates, in particular with regards to Sup35, are not understood. The yeast prion, URE3, is cured by overexpression of Hsp42, Btn2, Cur1, and Ydj1, proteins that are also involved in the clearance of non-amyloid aggregates. Interestingly, these proteins are localized to different cellular compartments. Ydj1 and Hsp42 are cytosolic proteins, whereas Cur1 is nuclear, Btn2 shuffles between the nucleus and the cytosol. The overexpression of Btn2, Hsp42, Cur1, or Ydj1 in URE3 yeast aggregates the URE3 seeds, which causes asymmetric segregation of the prion seeds during cell division. Overexpression of Btn2, Cur1 or Ydj1 causes the endogenous Hsp42 to form an aggregate, which binds the URE3 seeds. When Hsp42 is deleted, overexpression of Cur1 or Ydj1 still cures URE3, but at a significantly slower rate, whereas overexpression of Btn2 does not cure. Aggregation of the URE3 seeds by Hsp42 only occurred with full-length Hsp42, truncation of either the N-terminal prion domain or the intrinsically disorder domain of the Hsp42 eliminated aggregation of the Ure2 seeds. Our results suggest that the full-length Hsp42 acts as a scaffold that binds the prion seeds, thereby contributing to the curing of URE3 prion by asymmetric segregation of the seeds. We are also investigating the role of Hsp104 in curing prion and in aggregating huntingtin exon 1 fragments in yeasts. To maintain a steady state number of prion seeds from generation to generation, there must be constant propagation of the prion seeds, which is dependent on the severing activity of Hsp104. Paradoxically, Hsp104 overexpression as well as loss of Hsp104 activity cures the PSI+ prion. Two very different models of curing have been proposed: dissolution of the prion seeds by Hsp104 overexpression and curing by asymmetric segregation of the prion seeds. To better understand the mechanism of curing PSI+ by Hsp104 overexpression, we examined whether the rate of curing is affected by different members of the Hsp70 family. The rate of curing of PSI+ by Hsp104 overexpression is significantly faster in yeast expressing only Ssa1 than in yeast expressing only Ssa2 even though these homologs are 98% identical. At the same time, the number of PSI+ seeds are not significantly different whether the yeast are expressing Ssa1 or Ssa2. Since Ssa1/SSa2 affected the rate of curing by PSI+ by overexpression of Hsp104 without affecting seed number these results show that the mechanism of curing by Hsp104 overexpression is independent of the severing activity of Hsp104. In yeast, unlike in mammalian cells, the aggregation of huntingtin fragments with expanded polyglutamine repeat regions (HttpolyQ) has been found to be dependent on prion. This, in turn, makes HttpolyQ aggregation dependent on Hsp104, which is required to propagate the prion seeds. However, it is not clear whether Hsp104 has any other role in HttpolyQ aggregation. In the present study, we find that Hsp104 has a marked effect on the aggregation of HttQ103 fragments both in the presence and absence of prion. In the presence of both prion and Hsp104 the cells have numerous aggregates, but when Hsp104 is inactivated, the cells have a large HttQ103 aggregate formed from coalescence of the smaller aggregates. In the absence of both prion and Hsp104, there is no significant aggregation of HttQ103, but with active Hsp104 in the absence of prion, the cells slowly accumulate aggregates; presumably, the Hsp104 is amplifying aggregates from the small amount of spontaneous nucleation that is known to occur. After a week, the cells with Hsp104, but without prion, have numerous HttQ103 aggregates per cell and the biophysical properties of these aggregates are indistinguishable from the aggregates that form with both prion and Hsp104. These results show, that regardless of whether there is spontaneous or prion-templated nucleation of HttQ103, Hsp104 severs the initial HttQ103 aggregates, thus increasing their number and enhancing their transmission to the daughter cells. In contrast to HttQ103, HttQ103P, which has a polyproline region downstream of the polyglutamine region, requires prion to accumulate aggregates; aggregates do not accumulate in the presence of Hsp104 alone. Furthermore, the HttQ103P aggregates that are nucleated by prion are not severed by Hsp104, whether Hsp104 is present or not, these small aggregates coalesce into one large HttQ103P aggregate. We conclude that the ability of Hsp104 to sever either the spontaneously or prion-templated nuclei of HttQ103, but not the equivalent nuclei of HttQ103P, accounts for the differences in the aggregation properties of HttQ103 and HttQ103P.