In the present study, we reveal global differences in gene expression between yeast cells lacking two cytosolic HSP70s, SSA1 and SSA2, and the mild-heat-shocked wild-type using cDNA microarray technologoly.
Results from cDNA microarray analysis reveal that the stress-inducible protein genes, including molecular chaperones, were up-regulated in the ssa1/2 deletion mutant in a similar fashion as seen in the mild heat-shocked wild-type (Figs. 1A, 2A, and 5B). It is clear that thermotolerance is due to expression of these stress-inducible proteins. In the ssa1/2 deletion mutant, HSF1 suppressing growth rate of the ssa1/2 [15, 16] was expressed normally and its expression level was unchanged (data not shown). Several genes involved in the ubiquitin-proteasome protein degradation pathway were up-regulated in the ssa1/2 deletion mutant (Fig. 5D and Table 2). UBC4 [33, 34] was also up-regulated in the ssa1/2 deletion mutant, which is consistent with a previous report . UBC4/5 is necessary for binding between the substrates and Lys48 of ubiquitin that is a target of the 26 S proteasome , and for binding between the substrates and Lys63 of ubiquitin, that is not a target of the 26 S proteasome. In addition to UBC4, we found up-regulation of several proteasome genes (PRE1, RPN4, RPN12 and SCL1) in the ssa1/2 deletion mutant. PRE1 and SCL1 encode 20 S proteasome, and RPN4 and RPN12 encode 26 S proteasome . RPN4 (SON1) is a factor involved in ERAD (endoplasmic reticulum associated degradation) . All these genes are essential for degradation of the ubiquitinated proteins [36, 38]. RT-PCR data support the up-regulation of these proteasome genes by the deletion of SSA1/2 (Fig. 6A). Moreover, we confirmed that Pre1p and Rpn4p were up-regulated in the ssa1/2 deletion mutant at the translational level by immunoblotting (Fig. 7). This result provided further evidence that proteolytic degradation by proteasomes was stimulated by the deletion of SSA1/2. As shown in Fig. 8, more ubiquitinated proteins, especially with molecular weights less than 30 kDa, were detected in the ssa1/2 deletion mutant than in the wild-type. The deletion of UBP3 in ssa1/2 has been reported to lead to a significant increase in the number of the ubiquitinated proteins, mainly with molecular weights of more than 30 kDa . The expression level of UBP3 did not change in the ssa1/2 deletion mutant compared with the wild-type (data not shown). Therefore, the increase of ubiquitinated proteins in ssa1/2 is not caused by the deletion of UBP3. There are two possibilities for the increase of ubiquitinated proteins in the ssa1/2 deletion mutant. First, insufficiency of the UPR in the ssa1/2 deletion mutant may lead to activation of the ubiquitin-proteasome protein degradation system. Thus, ubiquitination of the target proteins increases and the expression of proteasome genes is induced. Second, some defect of deubiquitination occurs in the ssa1/2 deletion mutant, consequently leading to the accumulation of ubiquitinated proteins in the cell followed by cell death. Furthermore, proteolytic degradation by proteasomes is facilitated. On the other hand, we found that several genes encoding ribosomal proteins were up-regulated in the ssa1/2 deletion mutant (Figs. 5C, 6A, and Table 1), implying that protein synthesis is activated by the deletion of SSA1/2.
In case of the mild heat-shocked wild-type, genes involved in protein synthesis were significantly suppressed (Figs. 2B and 4B), and the proteasome genes up-regulated in the ssa1/2 deletion mutant did not show any change in their expression levels (Fig. 5D). Instead, some UPR genes (PDI1, DER1, ERO1 and KAR2) were up-regulated (Fig. 5D), implying that UPR occurs during mild heat-shock. The mechanism of UPR is known to induce the up-regulation of ER chaperones for refolding when unfolded proteins accumulate in the ER . In the ssa1/2 deletion mutant, the expression of three UPR genes (PDI1, DER1 and ERO1) remained unchanged (data not shown), and only KAR2 was up-regulated (Fig. 3).
Gasch et al. has reported the genome-wide expression analysis of yeast cells exposed to environmental changes . We compared our data on the mild heat shocked wild-type yeast cells with their data on the wild-type cells shifted to 37°C from 25°C. The stress-inducible protein genes up-regulated in the mild-heat shocked wild type (SSA3, SSA4, SSE2, CTT1, HSP26, HSP78, and HSP104) (Fig. 5B) are in common with the results obtained by Gasch et al. . In the category of "Protein fate", more than 70% of the up-regulated genes in our experiments are also in common with their results , even though there is a time lag with their experiments. However, DER1, one of the UPR genes, was not found to be up-regulated in their study during the entire the heat-shock treatment period . In contrast, DER1 was up-regulated in our experiments (Fig. 5D). This may be due to the difference in the temperature and time of heat-shock treatment. On the other hand, in the category of "Protein synthesis", ribosomal protein genes are significantly suppressed in their experiments , which is consistent with our data. From these comparisons, it can be said that our data on the mild heat-shocked wild-type is similar to that reported by Gasch et al.  in the categories of "Cell rescue, defense, and virulence", "Protein fate" and "Protein synthesis".
It is reasonable that UPR is activated and protein synthesis is suppressed in the mild heat-shocked wild-type. We speculate on the reasons as to why the genes involved in both protein degradation and protein synthesis are up-regulated in ssa1/2 deletion mutant. In the normal state, proteins are synthesized on the ribosome, followed by post-translational modifications in the ER or the Golgi apparatus to finally become mature and functional entitles. Schubert et al  showed that 30% of the de novo synthesized proteins are degraded before coming to maturity. Therefore, it can be reasoned that post-translational protein denaturation occurs moderately even under normal conditions. However, organisms have developed several mechanisms in their response to the denatured proteins. UPR is one of the ER quality control mechanisms . In addition, the refolding of denatured proteins is carried out by cytosolic chaperones [25, 42, 43], including SSA1/2 [20, 21]. It can be hypothesized that the deletion of SSA1/2 leads to the suppression of refolding, which is then followed by an accumulation of the denatured proteins in cells. The genes involved in proteolytic degradation may be up-regulated to remove such denatured proteins. However, if the ubiquitin-proteasome system keeps on degrading proteins, the depletion of the proteins essential for growth and development will occur. It is suggested that protein synthesis is activated to supply the proteins deleted by proteolytic degradation in the ssa1/2 deletion mutant. In the ssa1/2 deletion mutant, several hexose transporter genes (HXT2, HXT4, HXT6, HXT7), and the genes that belong to early part of glycolysis (GLK1, HXK1) were up-regulated (data not shown). The expression of these genes, involved in energy generation, may be required for sustaining the increased protein synthesis in the ssa1/2 deletion mutant. HXT genes up-regulated in the ssa1/2 deletion mutant are low-glucose dependent [44–46]. It is possible that the uptake of glucose is activated to generate energy, because energy is consumed by protein synthesis that is induced by the deletion of SSA1/2.
These results indicate that different mechanisms of the response to denatured proteins are employed between the ssa1/2 deletion mutant and the mild heat-shocked wild-type even though several up-regulated Hsps (molecular chaperones) are common between the ssa1/2 deletion mutant and the mild heat-shocked wild-type (Fig. 5B). When Hsp104p, Ydj1p (yeast Hsp40p), and Ssa1p exist together, their chaperone activities increase significantly . From this, it is suggested that the deletion of SSA1/2 induces the suppression of their chaperone activities. Recently, the cooperation of Hsp26p wih Hsp104p/Hsp70p/Hsp40p chaperone system on protein disaggregation in yeast was reported [47, 48]. Hsp26p co-aggregated with substrate is suggested to be a target of the Hsp104p/Hsp70p/Hsp40p chaperone system [47, 48]. Although Ssa1p is able to disaggregate the early Hsp26p-substrate complex (small soluble aggregates), Hsp104p is essential in refolding the late Hsp26p-substrate complex (big insoluble aggregates) [47, 48]. Moreover, excess or stoichiometric Hsp26p against denatured substrates is essential for effective refolding . In the ssa1/2 deletion mutant, an increase in the mRNA expression levels of HSP104 and HSP26 was seen (Fig. 5B). Although the refolding of denatured proteins is sure to succeed if HSP104/Hsp104p and HSP26/Hsp26p are highly expressed, it is a fact that the ubiquitin-proteasome degradation system is facilitated in the ssa1/2 deletion mutant. It can be speculated that as constitutive protein denaturation occurs, the ubiquitin-proteasome degradation system is required in addition to the chaperone refolding system in the ssa1/2 deletion mutant. Furthermore, there is a possibility that protein refolding by molecular chaperones and ubiquitin-proteasome protein degradation are related. In mammalian cells, the following model has been reported; denatured proteins are refolded by Hsp70-HSP40 chaperone-mediated maturation pathway under the treatment of Hsp90 inhibitor, and then denatured proteins are degraded by ubiquitin-proteasome . It is interesting to note that the ubiquitin-proteasome protein degradation system in yeast is induced when the chaperone function is inhibited by the deletion of SSA1/2. However, at present, our data are not sufficient to propose a similar model in yeast, and this remains a topic for future study.