Hsp70 and Hsp40 Chaperone Activities in the Cytoplasm and the Nucleus of Mammalian Cells*

The existence and function of a Hsp40-Hsp70 chaperone machinery in mammalian cells in vivo was investigated. The rate of heat inactivation of firefly luciferase transiently expressed in hamster O23 fibroblasts was analyzed in cells co-transfected with the gene encoding the human Hsp40 (Ohtsuka, K. (1993) Biochem. Biophys. Res. Commun. 197, 235–240), the human inducible Hsp70 (Hunt, C., and Morimoto, R. I. (1985)Proc. Natl. Acad. Sci. U. S. A. 82, 6455–6459), or a combination of both. Whereas the expression of human Hsp70 alone in hamster cells was sufficient for the protection of firefly luciferase during heat shock, expression of the human Hsp40 alone was not. Rather, this led to a small but significant increase in the heat sensitivity of luciferase. The expression of the human Hsp40 only led to heat protection when the human Hsp70 was expressed as well. Under such conditions the rate of luciferase reactivation from the heat-inactivated state was increased, but the rate of inactivation during heat shock was not affected. Using constructs that direct firefly luciferase either to the cytoplasm or to the nucleus (Michels, A. A., Nguyen, V.-T., Konings, A. W. T., Kampinga, H. H., and Bensaude, O. (1995) Eur. J. Biochem.234, 382–389), it was demonstrated that these chaperone functions are found in both compartments. Our data provide the first evidence on how the Hsp40/Hsp70 chaperone complex acts as heat protector in mammalian cells in vivo.

Cells that have been subjected to a priming heat shock become less susceptible to a following challenging heat shock. This transient thermotolerant state is accompanied by a temporary increase in the expression of heat shock proteins (HSPs). 1 There is a large body of evidence that it is indeed this increase that protects the cells against heat (1). The precise mechanism of single HSPs during in vivo heat protection is unknown, but overexpression of only one HSP (for example Hsp70) can be sufficient for thermoprotection in terms of clonogenic survival (2,3).
Being molecular chaperones, one of the mechanisms of protection of individual HSPs could be in preventing thermally denatured proteins from subsequent aggregation (4). Indeed, for Hsp70, indirect evidence (5) has suggested that its overexpression can protect cells from heat-induced protein aggregation in vivo. Such a function is consistent with the suggested in vitro chaperone capacity of the Hsp70 family (4). Hsp70 and the cognate Hsp70 (Hsc70) chaperones participate in many biological processes in which protein folding is involved. These include protein translocation, protein translation, protein assembly and disassembly, and protein degradation (4,6).
Chaperone functions have been investigated in vitro and in vivo using reporter proteins such as luciferase, ␤-galactosidase, citrate synthase, and dihydrofolate reductase. Overexpression of Hsp70 protects luciferase against thermal denaturation in vivo (7). On the other hand, neither overexpression nor deletion of the prokaryotic Hsp70, DnaK, affects heat-induced inactivation of the foreign reporter enzyme luciferase in Escherichia coli (8). Also, heat-induced luciferase aggregation within yeast mitochondria is not influenced by mutations in mitochondrial Hsp70 (9).
Several studies in prokaryotes have shown that the bacterial homologue of Hsp70, DnaK, is assisted by two co-chaperones, DnaJ and GrpE (6,10,11). DnaJ stimulates the ATPase function of DnaK (12) whereas GrpE enhances the ADP/ATP exchange on DnaK (10). DnaJ has a chaperone function of its own. Like DnaK, DnaJ is capable of binding to substrate proteins and can prevent luciferase from aggregation in vitro (8).
A bovine DnaJ homologue was found to reduce the amount of Hsc70 needed to refold heat-inactivated DNA polymerase in vitro (13). In human cells, Hsp40 (HDJ-1), a heat-inducible DnaJ homologue (14), was found to co-localize with Hsc/Hsp70 before and after heat shock (15). Hsp40 is able to assist Hsc70 in the de novo protein folding in in vitro translation systems (16). In vitro, Hsc70 binding to substrates prone to thermal aggregation was found to rely on the Hsp40-enhanced ATP hydrolysis (17). Together with Hsc70, Hsp40 can protect luciferase and rhodanese from, respectively, heat and chemically induced aggregation in vitro (17). This is suggested to be related to the finding that Hsp40 is capable of enhancing the reactivation of thermally denatured luciferase in vitro in a reaction in which bovine Hsc70 is included (17,18). This human DnaJ homologue is also capable of stimulating Hsp70-and Hsc70-assisted refolding of chemically denatured firefly luciferase and E. coli ␤-galactosidase (19,20).
The purpose of the current study was to examine the role of human Hsp70 and Hsp40 in the protection against and the recovery from heat-induced protein denaturation in mamma-lian cells. As a model system, firefly luciferase was expressed in mammalian cells. This foreign reporter protein is inactivated upon mild heat treatment of the mammalian cells and can be reactivated during recovery after stress (21). The rate of luciferase inactivation was found to be reduced in thermotolerant cells (22,23). Considering the heat-induced Hsp40/Hsp70 relocalization from the cytoplasm into the nucleus, recombinant luciferases localized either in the cytoplasm (cyt-luciferase) or in the nucleus (nuc-luciferase) were used as reporters of damage in either compartment (23). By coexpressing human Hsp70 (hu-Hsp70) and/or human Hsp40 (hu-Hsp40) with luciferase in O23 hamster cells, we examined whether these HSPs would function as chaperones on their own and/or whether they could cooperate in the heat protection of firefly luciferase in vivo.

EXPERIMENTAL PROCEDURES
Plasmids and Cloning Techniques-Plasmid pRSVLL/V (24), encoding for a cytoplasmic localized firefly luciferase, was kindly provided by Dr. S. Subramani, University of California, San Diego, CA. To direct luciferase into the nucleus, we recently described the construction of pRSVnlsLL/V (23).
Plasmids pCMV40 and pCMV70 are used to express the human Hsp40 or the human-inducible Hsp70, respectively. To construct plasmid pCMV40, a HindIII-BamHI fragment encoding the cDNA sequence of human Hsp40 (25)  (v/v) fetal calf serum (Life Technologies, Inc.). 1.5 ϫ 10 5 cells were seeded in a 25-cm 2 culture dish. They were transfected after 24 h with either 1 g of pRSVLL/V or pRSVnlsLL/V with or without (combinations of) 1 g of pCMV40 and pCMV70 following the standard calcium phosphate method (27). The plasmid quantity was kept equal to 10 g/25 cm 2 dish by the addition of plasmid pSP64 (Promega). One day after transfection, the transfected cells were treated with trypsin and distributed equally into cell culture tubes (Nunc); 40 tubes (1 ϫ 10 4 cells/tube) were seeded out of a 25-cm 2 culture dish, and 20 mM MOPS was added to the medium to increase its buffering capacity. Heat shocks were given 2 days after transfection. For inactivation experiments, the tubes were preincubated in a water bath at 37°C and transferred to a second water bath at the indicated temperatures Ϯ 0.1°C. The transfer took less than 3 s and was taken as time 0. For recovery experiments, the medium was replaced 30 min before heat shock by medium containing 20 g/ml cycloheximide to inhibit new protein synthesis. The inhibition of protein synthesis was checked by 35 S incorporation (data not shown). After various heat shocks, cells were allowed to recover at 37°C. For both inactivation and recovery experiments, the activity before heat shock was taken as 100%.
Immediately after treatment, cells were cooled to 4°C and lysed. The lysis procedure was initiated by a brief wash with ice-cold PBS and subsequent cell lysis in 500 l of buffer A (25 mM H 3 PO 4 /Tris, pH 7.8, 10 mM MgCl 2 , 1% (v/v) Triton X-100, 15% (v/v) glycerol, 1 mM EDTA) containing 0.5% (v/v) 2-mercaptoethanol. The lysates were kept frozen at Ϫ20°C before measurement of luciferase activity. Luciferase activities were measured in a Berthold Lumat 9501 for 10 s after the addition of substrates (1.25 mM ATP and 87 g/ml luciferin (Sigma) in buffer A).
Western Blotting-O23 cells in a 25-cm 2 dish were transfected with pCMV40 or pCMV70 or pCMV5 as a control and trypsinized 2 days after transfection. SDS (1% (w/v)) and 2-mercaptoethanol were added to the cells, and they were sonicated and subsequently heated for 5 min at 90°C. After electrophoresis through a 10% polyacrylamide SDS gel, the proteins were electrotransferred onto nitrocellulose (Schleicher & Schü ll).
Insolubilization Experiments-O23 cells in a 25-cm 2 dish were transfected with 10 g pRSVLL/V, pRSVnlsLL/V, pCMV40, or pCMV70. For determination of protein solubility, heated and unheated cells were lyzed in buffer A and fractionated by centrifugation at 12,000 ϫ g at 4°C for 15 min into supernatants and pellets. SDS (1%: v/v) and 2-mercaptoethanol (1%: v/v) were added to the supernatants. The pellets were dissolved in SDS (1%: v/v) and 2-mercaptoethanol (1%: v/v). The samples were analyzed on 10% polyacrylamide SDS gels and analyzed for the presence of luciferase, Hsp40 or/and Hsp70 by Western analysis.
Immunohistochemistry-Cells were transfected with 1 g of pCMV40 or pCMV70. One day after transfection, cells were plated on coverslips. One day after plating, the coverslips were processed for immunofluorescence at room temperature. The cells were washed with PBS, fixed with 3.7% (v/v) paraformaldehyde for 15 min, then washed sequentially 3 ϫ 5 min with PBS and incubated 15 min with 0.2% (v/v) Triton X-100 in PBS then 10 min with 100 mM glycine in PBS. Thereafter the coverslips were soaked for 30 min in 3% (w/v) BSA in PBS, and the cells were incubated for 1 h with a 1:150 dilution of either anti-Hsp70 (C92F3A-5, Stress-gen) or an anti-Hsp40 serum (28) in 0.1% (v/v) Tween-20 in PBS (Tween/PBS). The coverslips were washed in Tween/ PBS and incubated for 1 h with a 1:150 dilution of a Cy3-conjugated anti-mouse antibody (Amersham) (Hsp70) or a fluorescein-conjugated (FITC) anti-rabbit antibody (Nordic) (Hsp40). After 3 washes in Tween/ PBS, the coverslips were mounted with Vectashield (Vector Laboratories, Inc.). The fluorescence was detected by a Molecular Dynamics laser-scanning confocal microscope with a Zeiss Plan-Apochromat 63 ϫ/1.40 oil immersion objective. Image processing did not involve either background or contrast adjustments.

Thermal Denaturation of Luciferases in Vivo-
To express nuc-luciferase or cyt-luciferase, O23 hamster cells were transiently transfected with either pRSVnlsLL/V or pRSVLL/V, respectively. The cells were heated at various temperatures, and luciferase activity in the lysates was followed as a function of time (Fig. 1). The activity decayed following a first order reaction from 2 to 10 min heating at 40°C, 2 to 6 min at 42°C, or 2 to 4 min at 44°C. Upon heat shocks exceeding the abovementioned intervals, the inactivation kinetics departed from an exponential decay. Luciferase inactivation rate constants (k) were determined from the initial portion of the curves (Table I). When the temperature was elevated from 40 to 44°C, the inactivation rates (k) increased 8 -11-fold in control cells transfected with either nuc-luciferase or cyt-luciferase. As described before in murine cells (23), in O23 hamster cells nuc-luciferase was more rapidly heat-inactivated than cyt-luciferase at all temperatures tested.
Expression of the Human Hsp40 Sensitizes Luciferase Denaturation-Next, the effect of expression of the hu-Hsp40 on firefly luciferase thermostability was examined. O23 cells were transiently transfected with the pCMV40 expression vector, carrying the cDNA for the human DnaJ homologue (14). Ex- pression of the hu-Hsp40 was evaluated by Western blotting (Fig. 2). In lysates from unstressed, nontransfected cells or cells transfected with the pCMV5 empty vector, a weak Hsp40 band but no Hsp70 band was detected. In contrast, high amounts of both Hsp40 and Hsp70 were observed in untransfected cells that had been made thermotolerant by a priming heat shock followed by a recovery at 37°C. Lysates from O23 cells transfected with pCMV40 showed a substantial increase in Hsp40 expression levels. No induction of the hamster Hsp70 was seen, indicating that expression of the hu-Hsp40 did not induce a general stress response and that the observed increase in Hsp40 signal can be assigned to the expression of the hu-Hsp40 and not the endogenous hamster Hsp40.
In cells co-transfected with either cyt-luciferase or nuc-luciferase expression vectors and the pCMV40 expression vector, no protective effect of hu-Hsp40 on heat inactivation of luciferase was seen (Fig. 3). Rather, inactivation rates were higher in cells expressing the hu-Hsp40 (open circles) compared with control cells (closed circles). For better quantitative comparison of protein inactivation in the absence or presence of HSPs, a protection factor (PF) was calculated by dividing the rate constant determined in the presence of co-transfected HSP by the rate constant of the corresponding control (Table I). At all temperatures tested, the PF for the hu-Hsp40 was below 1, indicating an unexpected heat sensitization rather than a protecting effect of this HSP.
Expression of the Human Hsp70 Attenuates Luciferase Denaturation-Expression of the hu-Hsp70 using pCMV70 as expression vector was possible in hamster O23 fibroblasts without inducing a general stress response. As can be seen in Fig. 2, control cells did not show any expression of the heat-inducible hamster Hsp70, whereas the expression of Hsp70 was clearly visible in the pCMV70-transfected cells. This occurred without any concurrent changes in the expression level of the stressinducible hamster Hsp40.
In cells expressing the hu-Hsp70, both luciferases were protected against heat inactivation at all temperatures tested (Fig.  3, closed squares; Table I). The PF for hu-Hsp70 was higher than 1 under all heat shock temperatures and decreased with increasing heat shock temperatures. The level of protection of nuc-luciferase was lower than of cyt-luciferase at all temperatures tested (Table I).
Combined hu-Hsp70 and hu-Hsp40 Expression and Luciferase Inactivation-Next, it was tested whether the hu-Hsp40 might affect the protective action of the hu-Hsp70. Cells were co-transfected with either cyt-luciferase or nuc-luciferase together with both hu-Hsp70 and hu-Hsp40 (Fig. 3, open squares) and subsequently heated at different temperatures. It is clear from both  Table I that the co-transfection of the hu-Hsp40 with the hu-Hsp70 did not enhance (nor reduce) the protective action of the hu-Hsp70 alone.
Solubility of the Recombinant Proteins-Loss of solubility can TABLE I Thermal inactivation rates of nuclear and cytoplasmic luciferase and effects of hu-Hsp40 and hu-Hsp70 O23 cells were transiently transfected with pRSVnlsLL/V or pRSVLL/V and (combinations of) pCMV40 and pCMV70. They were heated at 40, 42, and 44°C, and inactivation rates (k) were determined from the log-linear part (2-10 min at 40°C, 2-6 min at 42°C, and 2-4 min at 44°C) of the curves (Fig. 3 and data not shown). The PFs were calculated by dividing the k from the inactivation curve of cells transfected with luciferase only by the k of the corresponding Hsp-transfected cells.  be used as an indicator of protein denaturation in addition to the loss of enzymatic activity. The solubility of recombinant luciferases, Hsp40 and Hsp70, in nondenaturing lysis buffer was examined. Lysates from cells expressing one of the recombinant proteins were fractionated by centrifugation at 12,000 ϫ g into a supernatant (S) and a pellet (P) containing the nuclei, the mitochondria, and part of the cytoskeleton. The distribution of the recombinant proteins in the fractions was probed by Western blotting (Fig. 4). Both luciferases, Hsp40 and Hsp70, were detected in the soluble fraction from unstressed control cells.
Markedly different results were obtained when lysates from heat-shocked (0, 5, 10, and 20 min 43°C) cells were fractionated. As described before (23) both luciferases were redistributed into the pellet fractions, the nuclear luciferase more rapidly than the cytoplasmic luciferase. Like the luciferases, Hsp40 could be detected in the insoluble fraction after heatshock. Overexpressed Hsp70 remained soluble under these conditions.
hu-Hsp40 and hu-Hsp70 Localization-Overexpression of hu-Hsp40 or hu-Hsp70 affected inactivation kinetics of both nuclear and cytoplasmic luciferases. Both HSPs have been reported to be localized predominantly in the cytoplasm before heat shock and to relocalize to the nucleus after heat shock (15,29). Therefore, the localization of transfected hu-Hsp40 and hu-Hsp70 was checked. Cells transfected with the appropriate transfection vectors were processed by indirect immunofluorescence and were examined by confocal microscopy (Fig. 5). In cells overexpressing the hu-Hsp40 or hu-Hsp70, the proteins were not exclusively located in the cytoplasm. Clear staining was observed in the nucleus of nonheat-shocked cells as well. Therefore, with the current model we cannot detect whether heat-induced translocation or/and (low) pre-heat shock levels of these HSPs are responsible for their effects on luciferase inactivation in the nuclear compartment.
Enhanced Recovery upon hu-Hsp70 and hu-Hsp40 Coexpression-It has been shown that a variety of mammalian cells expressing wild type, cytoplasmic, or nuclear luciferase were capable of partially reactivating these heat-inactivated reporter enzymes when allowed to recover at 37°C in the presence of a protein synthesis inhibitor (21,23). We first examined whether O23 cells were also capable of recovering luciferases from the heat-inactivated state. After mild heat shocks of 15 min at 42 or 43°C for example (Fig. 6), both cyt-luciferase and nuc-luciferase could be recovered. Recovery was less for nucluciferase (circles) compared with cyt-luciferase (squares), but this might relate to a higher initial inactivation of the nuclear luciferase. Measurements on (differences in) luciferase recovery were not hampered by protein degradation. Fig. 7 shows that the total amount of both luciferase proteins detected by Western blotting did not change during heating or up to 60 min thereafter.
Luciferase recovery was dependent on the severity of the inactivating heat shock as e.g. hardly no luciferase reactivation was seen in control cells within 60 min after a severe heat shock of 30 min at 44°C (Fig. 8A, closed circles). However, in cells expressing the hu-Hsp70, the recovery of both cytoplasmic and nuclear luciferase was strongly enhanced (closed squares). In contrast, in cells expressing the hu-Hsp40 alone, no recovery was observed within 60 min (open circles). The recovery in the hu-Hsp40-expressing cells was even reduced compared with the small recovery that could be observed when the control cells were allowed to recover 4 h after this severe stress (Fig. 8B). It should be kept in mind that hu-Hsp70 expression markedly protected luciferase against heat inactivation, whereas in hu-Hsp40-expressing cells, the initial inactivation was found to be higher.
When the hu-Hsp40 was expressed together with the hu-Hsp70 (open squares), the recovery of the cytoplasmic luciferase from a heat shock of 30 min at 44°C was the same as with the hu-Hsp70 alone (closed squares) (Fig. 8A, CYT). Intriguingly, this was not the case for the nuclear luciferase. In this case, the recovery of heat-inactivated luciferase was enhanced when the hu-Hsp40 was coexpressed with the hu-Hsp70 (open squares) compared with expression of the hu-Hsp70 alone (closed squares) (Fig. 8A, NUC). This effect was not due to differences in the extent of heat inactivation of the nuclear luciferase as hu-Hsp40 did not affect hu-Hsp70-mediated luciferase protection ( Fig. 1 and Table I). Thus, after a heat shock of 30 min at 44°C, the recovery of the nuc-luciferase but not the cyt-luciferase in the presence of the hu-Hsp70 was enhanced by coexpression of the hu-Hsp40.
In vitro studies have shown that the level of Hsp70-mediated recovery of luciferase from its denatured state is highly dependent on the stoichiometry of Hsp70 and luciferase (30). We therefore tested whether the coexpression of hu-Hsp40 might have changed the Hsp70/luciferase ratio in the cell. Cells were transfected with either luciferase and pCMV70 alone or with luciferase and both pCMV70 and pCMV40, and lysates containing equal luciferase activities were probed for Hsp70 levels (Fig. 9). The level of Hsp70 per active luciferase protein was found to be independent of the absence or presence of cotransfected hu-Hsp40.
Initial Inactivation and Luciferase Recovery upon hu-Hsp70 and hu-Hsp40 Coexpression-Finally, we addressed the question why the enhancement of hu-Hsp70 functions in the presence of the hu-Hsp40 during recovery was only observed for the nuclear luciferase and not for its cytoplasmic counterpart. So the importance of differences in initial inactivation on the recovery in the presence of hu-Hsp70 and hu-Hsp40 was investigated. When O23 cells expressing the hu-Hsp70 with or without the hu-Hsp40 were heated for 30 min at 42°C, the nucluciferase was inactivated to a level comparable to cytluciferase in cells heated at 44°C (Table II). Under these conditions of similar inactivation, the expression of hu-Hsp40 in cells co-transfected with hu-Hsp70 enhanced the recovery of the nuc-luciferase (Fig. 10A) but not that of the cyt-luciferase (Fig. 8A, CYT). But for conditions of more severe levels of similar initial inactivation (30 min at 46°C for the cyt-luciferase compared with 30 min at 44°C for the nuc-luciferase (Table  II)), expression of the hu-Hsp40 not only enhanced the recovery of the nuc-luciferase (Fig. 8A, NUC) but also that of the cytluciferase (Fig. 10F). Thus, in the presence of hu-Hsp70, hu-Hsp40 enhanced the recovery of both luciferases, but this enhancement was more pronounced when the inactivating heatshock, and thus the initial inactivation, was more severe. DISCUSSION In the current study it was found that cytoplasmic and nuclear-localized firefly luciferases were protected against heat FIG. 7. Luciferase degradation during heat shock and recovery. O23 cells transiently transfected with pRSVnlsLL/V (NUC) or pRSVLL/V (CYT) were heated for 30 min at 44°C and allowed to recover at 37°C for 0 -60 min. Control cells were kept at 37°C. Cycloheximide (20 g/ml) was present throughout the treatment to inhibit new protein synthesis. Cell lysates were loaded on 10% polyacrylamide gels and transferred to nitrocellulose. These were probed with an antiluciferase antibody. Upper lanes are for nuc-luciferase, bottom lanes are for cyt-luciferase.  9. Effect of coexpression of Hsp40 on the Hsp70 to luciferase ratios in cells. O23 cells were transiently transfected with pRSVLL/V and only pCMV70 (pCMV70) or the combination of pCMV70 and pCMV40 (pCMV70 ϩ pCMV40). Luciferase activity was measured, and cell lysates were loaded on 10% polyacrylamide gels on the basis of equal luciferase activities (relative light units (RLU)). These were probed for Hsp70 levels by Western blotting using an anti-Hsp70 antibody.
inactivation in O23 hamster cells expressing hu-Hsp70 and slightly heat-sensitized in cells expressing hu-Hsp40. The recovery of both luciferases was enhanced in cells expressing hu-Hsp70 and decreased in cells expressing hu-Hsp40. Although coexpression of both hu-Hsp70 and hu-Hsp40 lead to the same protection as hu-Hsp70 alone, recovery of both luciferases in the presence of hu-Hsp70 was enhanced when hu-Hsp40 was coexpressed. As such, these two HSPs might contribute to the increase in protein thermostability and to the accelerated recovery from protein damage found in thermotolerant cells (22,23,31,32).
Hsp70-mediated Protection against Denaturation-Expression of the hu-Hsp70 yielded a protective effect on luciferase both when localized in the cytoplasm and in the nucleus. The magnitude of the Hsp70-mediated protection decreased with increasing temperatures. The more sensitive nuclear luciferase (Ref. 23 and this report) is less protected by Hsp70 than the cytoplasmic luciferase. Thus, the degree of protection by Hsp70 decreases with increasing protein denaturation rates.
Endogenous proteins like the insulin receptor and DNA topoisomerase I are protected from heat inactivation in thermotolerant cells (33,34). In both cases, heat stress induces the formation of complexes between Hsc70 (a constitutively expressed Hsp70 homologue) and the corresponding protein (35,36). The existence of such complexes suggests that Hsc70 may play a role in protecting proteins from heat inactivation in vivo.
Although we do not have any direct biochemical evidence of a co-association, our fractionation studies indicate that complexes between Hsp70 and luciferase might have kept the luciferase in an active form. During heat shock, the denaturation status of a protein depends upon the balance of a heatinduced protein denaturing force and a protein renaturation force. Increasing the levels of Hsp70 would enhance its probability to bind to a heat-inactivated protein. As such, this might inhibit irreversible hydrophobic protein interactions and aggregation and trap denatured intermediates that are closer to the active form.
Expression of the hu-Hsp70 caused no detectable general stress response. Therefore, an increase in Hsp70 per se might be sufficient in protecting luciferase from heat inactivation (Ref. 7 and this work). Such effect might be extended to other proteins as well. Indeed, constitutive expression of hu-Hsp70 in Rat-1 cells reduces the extent of heat-induced aggregation of endogenous proteins (5,37). The in vivo protection against heat-induced protein damage observed in rodent cells expressing the hu-Hsp70 was seen for temperatures up to 46°C (Refs. 5 and 37 and this report). Therefore, in vivo hu-Hsp70 can function at high hyperthermic temperatures as a chaperone. This is consistent with in vitro data, showing that DnaK/Hsc70 can protect RNA polymerase and DNA polymerase against heat inactivation at 45°C (13,38).
Enhanced Luciferase Recovery after Heat Shock upon Expression of hu-Hsp70 -The enhanced recovery after heat shock upon expression of hu-Hsp70 may be related to the attenuation of heat damage associated with hu-Hsp70 expression. The rate of luciferase recovery decreased upon increasing stress severity. Protein denaturation intermediates may have different capacities to refold; with increasing stress, less refoldable intermediates will be favored. Freeman and Morimoto (20) found that Hsc/Hsp70 is unable to mediate refolding of chemically denatured ␤-galactosidase at temperatures above 41°C in vitro. Yet their data also indicated that Hsc/Hsp70 is able to prevent protein aggregation at the higher temperature in vitro. By interacting with partially denatured proteins, Hsp70 might prevent them from further unfolding into an irreversible, nonrefoldable state.
Hsp40-mediated Sensitization for Heat Shock-Overexpression of hu-Hsp40 was found to sensitize luciferase for heat inactivation rather than to protect. An elevated level of Hsp40 might have trapped unfolded luciferase and overridden its binding to the endogenous Hsc70. Consistent with this hypothesis, we observed that hu-Hsp40 partially co-fractionated with heat-insolubilized proteins. Indeed, the bacterial DnaJ associates with the unfolded, heat-denatured luciferase and prevents it from aggregating in vitro (8). However, although related, DnaJ and Hsp40 show distinct properties. Minami et al. (17) did not find a stable binding of hu-Hsp40 to unfolded luciferase. As an alternative hypothesis, Hsp40 overexpression might prevent the endogenous Hsp70 from protecting the luciferase. hu-Hsp40 was found capable of enhancing the ATPase activity of Hsc/Hsp70 (17). Although stimulation of the ATPase activity of DnaK by DnaJ can lead to a more stable substrate-chaperone  complex (39,40), high (imbalanced) ratios of Hsp40/Hsp70 might change the endogenous hamster Hsc70 ATP/ADP status (before and/or during heat shock) so that a premature dissociation of unfolded luciferase from the endogenous hamster Hsc70 occurs during heat shock. Hsp40 overexpression may also result in an inappropriate association that would titrate out one of the other, numerous components of the chaperone machinery interacting with Hsp70/Hsp40 (4,41). Such titration would impair the chaperone machinery and result in heat sensitization. However, the involvement of an imbalanced Hsp40/Hsp70 ratio is supported by the fact that coexpression of Hsp70 along with Hsp40 eliminated the heat-sensitizing effects of hu-Hsp40 expression.
Hsp40 Overexpression Decreases Recovery-Expression of hu-Hsp40 reduced the extent of luciferase recovery after heat shock. This finding seems to contrast the previously reported DnaJ requirement for luciferase reactivation after heat shock (8). However, in their recovery experiments, DnaK was always included. As mentioned above, when expressing hu-Hsp40 without hu-Hsp70, an imbalance of Hsp40 and the endogenous Hsc70 might develop. Indeed, high ratios of DnaJ/DnaK or of the yeast DnaJ homologue Ydj1 with Hsp70 have been demonstrated to decrease luciferase refolding in vitro (40,42).
Hsp40 Cooperates with Hsp70 for an Accelerated Recovery-The most intriguing observation of the current study was that overexpression of hu-Hsp40 markedly accelerated the recovery of luciferase when hu-Hsp70 was expressed as well. In transient transfection assays, it is not possible to quantify the absolute levels of Hsp70 and Hsp40 per cell. However, Western analysis revealed that co-tranfection of Hsp40 and Hsp70 (with luciferase) did not lead to significant changes in the relative Hsp70:luciferase ratio compared with cells transfected with Hsp70 alone (and luciferase). Therefore, the enhancement of luciferase renaturation in cells expressing both Hsp40 and Hsp70 likely cannot be explained by alterations in the Hsp70: luciferase ratio.
The capacity of hu-Hsp40 to enhance hu-Hsp70-mediated luciferase reactivation was more prominent for the more severe heat treatments under conditions where Hsp40 might become limiting. The enhancement of the hu-Hsp70-mediated reactivation by expression of hu-Hsp40 occurred after lower inactivation temperatures for the nuclear luciferase (which is the most heat sensitive) than for the cytoplasmic luciferase. Because lower levels of Hsp40 are found in the nucleus, this chaperone might become limiting in this compartment under milder stress.
The accelerating effect of the hu-Hsp40, when coexpressed with the hu-Hsp70, on recovery of luciferase was not ascribable to an enhancement of the Hsp70-mediated protection against initial inactivation. We only find accelerated recovery after severe heat treatments. Therefore, by trapping or through Hsp70 stimulation, Hsp40 may have altered the severity of the damage to the luciferase, which is not directly reflected at the level of inactivation. The less severely damaged luciferase is next more rapidly recovered. Alternatively, the function of Hsp40 to stimulate the ATPase of Hsp70 during the heat shock may not be essential for thermoprotection but may only be relevant for recovery. Such would be consistent with the findings that Rat-1 cells expressing mutant hu-Hsp70 deleted of its ATP binding domain are resistant in terms of heat killing (43).
Although we have no definite proof that complexes between Hsp70, Hsp40, and (heat-denatured) luciferase do exist in vivo, our data support the suggestion from in vitro studies that also in intact mammalian cells, the hu-Hsp40 may act as a cochaperone of Hsp70 in assisting the refolding of luciferase from the heat-denatured state. Such cooperation has clearly been demonstrated in vitro (17). Consistent with our data, recently a Mdj1p/mtHsp70 (mitochondrial Hsp40/Hsp70 homologues) collaboration in thermoprotection has been observed within yeast mitochondria where Hsp70-luciferase complexes decreased in Mdj1p-deleted strains (9).