Description

The 90 kDa heat shock protein, Hsp90, is an evolutionarily conserved and highly expressed molecular chaperone, constituting 1–2% of cellular proteins under non-stressed conditions (Csermely et al. 1998). The abundance of Hsp90 is further increased by heat stress in virtually all cells (Lindquist and Craig 1988).

In fission yeast, Hsp90 (also known as Swo1 or Git10) is essential and contributes to various cellular processes, including glucose response (Hoffman and Winston 1991), cell cycle control (Aligue et al. 1994), cytokinesis (Santino et al. 2012), and heterochromatin assembly (Okazaki et al. 2018). The levels of Hsp90 protein in fission yeast increase upon heat stress and continuously rise at least for 2 hours (Aligue et al. 1994). A recent study has revealed that Hsp90 accumulates in a few granules in the cytoplasm under heat stress conditions (Cabrera et al. 2020), yet only limited information on the spatiotemporal distribution of Hsp90 is currently available.

To visualize the localization of Hsp90, we generated a yeast strain that expresses C-terminally GFP-tagged Hsp90 under its own promoter at the endogenous genomic locus. At the normal growth temperature (27˚C), Hsp90 was detected largely in the cytoplasm except for vacuoles and relatively less distributed in the nucleus (Fig. 1A). Under the high-heat stress condition (HHS: 45˚C), Hsp90 was still mostly localized in the cytoplasm, yet formed several cytoplasmic granules (Fig. 1B-D). The number of Hsp90 granules appeared to increase after 10 minutes of HHS (4~9/cell) as compared to 5 minutes of HHS (2~5/cell) (Fig.1B and C), although those numbers might be undercounted due to the indistinct nature of granules. Quantification of the cells harboring the Hsp90 granules revealed that Hsp90 granules are generated in most cells within 5 minutes after HHS and disappeared within 60 minutes upon HHS (Fig. 1G).

To characterize Hsp90 granules, we examined whether they co-localize with the stress-induced membrane-less organelles termed stress granules (SGs) by co-expressing a representative SG marker Pabp tagged with tdTomato. We found that Hsp90 granules overlap with Pabp (Fig.1B and C, arrows; Fig. 1H), although some granules do not overlap clearly (Fig. 1B and C, arrowheads). Notably, unlike Hsp90 granules, Pabp granules remain even after 60 minutes of HHS (Fig. 1E and F), thus the physical overlap between Hsp90 granules and Pabp granules is transient.

We next observed the localization of Hsp90 under a mild-heat stress condition (MHS: 37˚C), below the temperature at which SGs are generated. We found that MHS also induced the granule formation of Hsp90 in most cells within 5 minutes (Fig. 1J). In contrast to the HHS condition, Hsp90 granules did not disappear. Rather, both the size and the intensity of the Hsp90 granule fluorescence continuously increased at MHS (Fig. 1J-O), although the number of Hsp90 granules was less (2~3/cell). These findings indicate that the localization of Hsp90 is regulated differentially by temperatures.

Finally, our observation also revealed that both mild- and high-heat stress cause the nuclear accumulation of Hsp90 within 60 minutes (Fig. 1E-F, M-N, P). The location of the nucleus was confirmed by the nuclear envelope protein Cut11 tagged with mCherry (Fig. 1Q). Future studies will elucidate the mechanism and the biological significance of the nuclear translocation of Hsp90 upon heat stress. Given the involvement of Hsp90 as a part of the platform for building SGs under severe heat stress (Cabrera et al. 2020) as well as the heterogeneity of SGs upon various temperatures (Kanda et al. 2021), our findings will serve as a basis for understanding the regulation of thermo-sensitive aggregations.