INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

CORTISOL, THE PREDOMINANT glucocorticoid in vertebrates, including teleost fish, plays an important role in the stress response process, especially in allowing animals to regain homeostasis (21). The major pathway for cortisol signaling is genomic and is mediated by the cytoplasmic glucocorticoid receptor (GR), which functions as a transcription factor on ligand binding (1). A well-studied metabolic action of cortisol during stress involves GR-induced glucose production (21, 26) and this response is mediated by the synthesis of key proteins, including the gluconeogenic enzyme phosphoenolpyruvate carboxykinase (9). Consequently, physiological and/or pharmacological conditions that lower GR capacity may result in decreased metabolic responses to stress, including the heat-shock response (4).

The heat-shock response involves the synthesis of a suite of highly conserved heat-shock proteins (hsps) as part of the cellular stress response process to cope with the insult (17). The 70-kDa family of hsps (hsp70) are abundant and the most widely studied of the hsps and are known to have a number of important chaperoning functions including aiding in the folding of new proteins, refolding of incorrectly folded proteins, reducing protein aggregates, presenting proteins in a conformation suitable for degradation by the proteasome, and presenting steroid receptors in a ligand-binding conformation (7). The induction of hsp70 in response to stressors is thought to be critical to prevent proteotoxicity and enhance cell survival (10) and is perhaps the reason for the preferential synthesis of hsps even at the expense of other cellular proteins (4).

Recent studies allude to a possible link between the stress-induced cortisol response and the cellular heat-shock response. Chronic glucocorticoid treatment in vivo decreased the heat shock-induced hsp70 expression in two species of teleost fish liver (3) and rat brain slices (2). Also, cortisol attenuated the heat shock-induced hsp90 mRNA accumulation in trout hepatocytes in primary culture (22). The mechanism(s) for this attenuation is not known. A recent study using transformed cell lines expressing a human hsp70 promoter provided evidence for a GR-mediated inhibition of heat-shock factor 1 (HSF1) as a possible mechanism for the attenuated heat-shock response (27). Considered together, these results argue for a negative impact of stressor-mediated glucocorticoid stimulation on the heat-shock response. As elevated levels of glucocorticoid, either due to stressor exposure and/or pathological conditions, cause GR downregulation and restrict GR signaling (28), it appears likely that GR dynamics may be involved in the attenuated hsp70 response with cortisol (3), but this has not been tested before.

We set out to characterize the role of chronic glucocorticoid stimulation on hsp70 response using a physiologically relevant model system, the trout hepatocytes in primary culture (4, 22). The role of cortisol on hsp70 response was examined by measuring protein synthesis and degradation by a combination of ³⁵S-labeling and immunodetection, using trout-specific hsp70 and GR antibodies. Proteasomal inhibitors, lactacystin and MG-132 (14, 15), were used specifically to determine the role of the proteasome on GR and hsp70 turnover. Our results implicate proteasome-mediated GR degradation as a possible mechanism for the attenuated hsp70 response with cortisol in heat-shocked trout hepatocytes.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Primary cell culture. Rainbow trout (Oncorhynchus mykiss) were obtained from Rainbow Springs Trout Farm (Thamesford, Ontario, Canada) and were maintained at 12 ± 1°C on a 12:12-h light/dark cycle. Trout were acclimated for at least 2 wk before experimentation and were fed once daily to satiety (3 point sinking food; Martin Mills, Elmira, Ontario, Canada). Rainbow trout hepatocytes were isolated using collagenase perfusion according to established protocols (22) and the cells were plated in 6-well Primaria plates (Becton Dickinson Labware) at a density of 1.5 × 10⁶ cells/well (0.75 × 10⁶ cells/ml) in L15 media (Sigma, St. Louis, MO) supplemented with 0, 100, or 1,000 ng/ml cortisol (hydrocortisone, Sigma). The hepatocytes were maintained at 13°C for at least 24 h before heat-shock exposures. The cells were heat shocked at 28°C (+15°C) for 1 h and then allowed to recover at 13°C. Individual experimental details including the timing of heat shock and recovery from heat shock are included in each figure legend. Cell viability was assessed by measuring lactate dehydrogenase (LDH) leakage and there was no significant difference in LDH leakage between the control and heat-shocked hepatocytes (4). All animal procedures, including the maintenance of animals and experimental protocols, were according to the guidelines of the Canadian Council for Animal Care and approved by the Office of Research Ethics, University of Waterloo.

³⁵S-labeling of proteins. The cells were starved of methionine by changing to methionine-free RPMI media (ICN, Costa Mesa, CA) 2 h before ³⁵S-labeling. Just before the labeling, the volume in each well was reduced from 2 to 1 ml. Cells were labeled with [³⁵S]methionine (1,175 Ci/mmol, 40 µCi/treatment, ICN) at 13° for 1 h and the ³⁵S incubation was stopped by either harvesting the cells or replacing the media with 2 ml fresh L15 media. Appropriate cortisol concentrations (0, 100, 1,000 ng/ml) were maintained throughout. The cell pellets, obtained by centrifugation at 13,000 g for 20 s, were lysed, and the proteins were solubilized by vortexing and then boiling for 5 min in Laemmli sample buffer before storage at 20°C.

Proteasomal pathway studies. Hepatocytes in primary culture were heat shocked (+15°C for 1 h) and the proteasomal inhibitors, lactacystin (10 µM, Calbiochem, San Diego, CA) or MG-132 (50 µM, Calbiochem), were added to each well. Cells were incubated an additional 24 h (13°C) with the inhibitors, after which the cells were pelleted and stored frozen (70°) for immunodetection. The MG-132 (carbobenzoxyl-leucinyl-leucinyl-leucinal) inhibits cleavage of hydrophobic or acidic substrates and, therefore, is a general inhibitor of the proteasome (20), whereas lactacystin, a naturally occurring bacterial compound, appears to be substrate specific (6).

SDS-PAGE and Western blotting. The sample protein concentration was determined by the bicinchoninic acid method (Pierce Chemical) using BSA as the standard. The samples were analyzed on 8 or 10% polyacrylamide gels using the discontinuous buffer system of Laemmli (13) with 30 µg total protein loaded per lane. Gels were stained with BioSafe Coomassie stain (BioRad, Mississauga, Ontario, Canada), dried, and the radioactive incorporation was visualized and quantified with a Molecular Dynamics Storm 860 phosphorimager. For Western blotting, the gels were transferred (20 V for 30 min) onto nitrocellulose membranes (BioRad) with a SemiDry Transfer Unit (BioRad) using transfer buffer consisting of 25 mM Tris pH 8.3, 192 mM glycine, and 20% (vol/vol) methanol. Blots were blocked (60 min) with 5% skim milk in TBS-t [20 mM Tris pH 7.5, 300 mM NaCl, 0.1% (vol/vol) Tween 20] with 0.02% sodium azide. Primary and secondary antibodies were diluted in the blocking solution to the appropriate concentrations as indicated. For hsp70, a polyclonal rainbow trout gonadal RTG-2 antibody was used at 1:3,000 dilution (E. P. M. Candido/G. K. Iwama, University of British Columbia, Vancouver, Canada). This antibody recognized the total hsp70 (hsp70 and hsc70) in rainbow trout tissues (8, 25). For GR, a polyclonal trout GR antibody was used at 1:1,500 dilution [B. Ducouret, University of Rennes, Rennes, France (23)]. Both trout hsp70 and GR antibodies were raised in rabbits and the secondary antibody was alkaline phosphatase-conjugated goat anti-rabbit (1:3,000, BioRad) antibody. The blots were incubated in primary antibody (60 min) at room temperature, washed (3 × 5 min) with TBS-t, incubated with secondary antibody (60 min), washed with TBS-t (2 × 5 min), and finally washed with TBS (1 × 10 min). The bands were visualized with nitroblue tetrazolium (0.033% wt/vol) and 5-bromo-4-chloro-3-indolyl phosphate p-toluidine salt (0.017% wt/vol, Fisher Scientific). Prestained low range molecular weight markers (BioRad; Phosphorylase B 112 kDa, BSA 81 kDa, ovalbumin 49.9 kDa, carbonic anhydrase 36.2 kDa, soybean trypsin inhibitor 29.9 kDa, lysozyme 21.3 kDa) were run on all gels. The protein bands were scanned, and the band intensities were quantified using the AlphaEase computer program (AlphaEase Innovatech).

Statistical analyses. Data are presented as percent change from control (means ± SE) and were compared using either one-way ANOVA (cortisol treatments or time post-heat shock) or two-way ANOVA (cortisol treatment and proteasomal inhibitors or time post-heat shock as independent variables). The percentages were transformed for homogeneity of variance although non-transformed values are shown in the figures. Where P indicated significance (<0.05), means were compared using the post hoc Student-Newman-Keuls multiple comparison test.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cortisol lowers hsp70 accumulation in trout hepatocytes in primary culture. Heat shock (+15°C for 1 h) resulted in a preferential incorporation of [³⁵S]methionine into a 70-kDa protein after 2 and 24 h of recovery from heat shock (Fig. 1A). This newly synthesized protein was confirmed to be the inducible hsp70 using a trout-specific total hsp70 antibody (Fig. 1B). The weak 70-kDa band seen in the absence of heat shock is hsc70 because the antibody cross-reacts with both the constitutive and the inducible forms of hsp70 (Fig. 1B). However, the lack of any change in hsc70 expression with heat shock, detected using a trout-specific hsc70 antibody (data not shown), further confirmed that the significantly higher hsp70 content at 6 and 26 h after heat shock was indeed the inducible hsp70 (Fig. 1C). The synthesis of [³⁵S]hsp70 was lower with cortisol at all time points tested post-heat shock (Fig. 2A) and this was reflected in the significantly lower hsp70 accumulation with cortisol after heat shock compared with the control cells (Fig. 2, B and C).

View larger version (31K): [in this window] [in a new window]   Fig. 1.   Effect of heat shock (HS) on heat-shock protein 70 (hsp70) synthesis. Rainbow trout hepatocytes were freshly isolated and plated 24 h in L15 medium (13°C). Cells were then incubated in the absence or presence of HS (+15°C; 1 h) and then 35S labeled for 1 h at either 2 or 24 h post-HS. A: representative radioactive phosphorimage of a 10% gel showing 35S incorporation. B: a representative Western blot of hsp70 expression, detected using a trout-specific total hsp70 antibody, either in the absence of HS (lane 1) or 24 h after heat shock (lane 2). hsc70, Constitutive heat-shock protein 70. C: hsp70 protein accumulation after heat shock; band intensities were quantified and expressed as percent (100%) of control (absence of HS); values represent means ± SE (n = 3 separate fish); symbols with the same letters are not statistically significant (P < 0.05, 1-way ANOVA).

View larger version (32K): [in this window] [in a new window]   Fig. 2.   Cortisol decreases HS-induced hsp70 accumulation. Hepatocytes were incubated with cortisol (0, 100, or 1,000 ng/ml) and 24 h later cells were given an HS (+15°C: 1 h) and allowed to recover for 24 h. Hepatocytes were 35S labeled (1 h) 2 or 24 h after HS and proteins were separated by 10% SDS-PAGE. A: representative radioactive phosphorimage of a 10% gel showing 35S incorporation. B: representative Western blot of hsp70 expression detected using a trout-specific total hsp70 antibody. C: bar graph showing the effect of cortisol on hsp70 protein expression at 2, 6, and 26 h post-HS; values are shown as change from 100% control (0 ng/ml cortisol) and represent means ± SE (n = 3 fish); 2-way ANOVA showed significant cortisol effect [100 (crosshatched) and 1,000 ng/ml (solid bar) cortisol was significantly lower than control (open bar)] (P < 0.05; 2-way ANOVA), but there were no significant effects of either time or interaction of cortisol and time on hsp70 expression.

Cortisol does not affect hsp70 breakdown in trout hepatocytes. We examined whether the lower hsp70 response with cortisol was due to increased hsp70 degradation by ³⁵S pulse chasing of newly synthesized hsp70 protein. Preliminary studies established ³⁵S incorporation to be higher at 2 h relative to 0 and 1 h after heat shock. Therefore, hepatocytes were labeled with [³⁵S]methionine at 2 h and the decay of [³⁵S]methionine was determined at 0, 4, and 24 h later. The newly synthesized [³⁵S]hsp70 dropped significantly by ~30% and ~45% at 4 and 24 h, respectively, compared with the 0 h value (Fig. 3, A and B). There was no significant difference in the [³⁵S]hsp70 decay at 24 h compared with 4-h postheat shock (Fig. 3B). Cortisol treatment had no significant impact on hsp70 decay in the present study (Fig. 3C).

View larger version (29K): [in this window] [in a new window]   Fig. 3.   Effect of cortisol on hsp70 protein degradation. Hepatocytes were incubated with cortisol (0, 100, 1,000 ng/ml), heat shocked (+15°C; 1 h) and then 35S labeled (13°C for 1 h) 2 h after HS. After 35S labeling, the radioactive media were replaced with fresh media (with appropriate cortisol concentrations) and cells were harvested 0, 4, and 24 h later. A: representative radioactive phosphorimage of [35S]hsp70 decay. B: temporal changes in band intensities of control hepatocytes were quantified and shown as percentage change from 0 time; values represent means ± SE (n = 3 fish); symbols with the same letters are not statistically significant (P < 0.05, 1-way ANOVA). C: bar graph showing the effect of 100 (crosshatched) or 1,000 ng/ml (solid bar) cortisol on hsp70 decay at 0, 4, and 24 h post-35S labeling; values are shown as change from 100% control (0 ng/ml cortisol; open bar) and represent means ± SE (n = 3 fish); 2-way ANOVA showed no significant effect of cortisol, time after labeling, or interaction of cortisol and time on hsp70 decay.

Cortisol downregulates GR in trout hepatocytes. As cortisol affects hsp70 synthesis, we examined whether GR signaling may be a factor in the attenuated heat shock response. Cortisol treatment significantly decreased GR protein expression in trout hepatocytes in the absence of heat shock (79 and 81% of control for 100 and 1,000 ng/ml cortisol, respectively). Cortisol also significantly decreased GR protein expression in heat-shocked trout hepatocytes (Fig. 4, A and B). Proteasomal inhibition by lactacystin and MG-132 significantly increased GR expression either in the absence or presence of cortisol (Fig. 4C), implicating a role for the proteasome in the GR degradation process in trout hepatocytes.

View larger version (30K): [in this window] [in a new window]   Fig. 4.   Glucocorticoid receptor (GR) downregulation is mediated by the proteasome. Hepatocytes were incubated with cortisol (0, 100, or 1,000 ng/ml), heat shocked (+15°C; 1 h), and then DMSO vehicle (control; open bar), 10 µM lactacystin (crosshatched), or 50 µM MG-132 (solid bar) was added to each well and sampled 24 h after the addition of inhibitors. A: representative Western blot of GR expression with cortisol and proteasomal inhibitors in heat-shocked hepatocytes; blots were probed using a trout-specific GR antibody. B: bar graph showing the effect of cortisol on GR expression in trout hepatocytes; values are shown as change from 100% control (0 ng/ml cortisol) and represent means + SE (n = 3 fish); bars with the same letters are not significantly different (P < 0.05, 1-way ANOVA). C: bar graph showing the effect of proteasomal inhibitors on GR expression either in the absence or presence of cortisol (100 or 1,000 ng/ml); values are shown as change from 100% control (absence of proteasomal inhibitor; open bar) and represent means + SE (n = 3 fish); 2-way ANOVA showed significant inhibitor effect [lactacystin (crosshatched) and MG-132 (solid bar) significantly increased GR expression compared with the control (open bar)] (P < 0.05; 2-way ANOVA), but there were no significant effects of either cortisol or interaction of cortisol and inhibitors on GR expression.

Proteasome is involved in the hsp70 response by cortisol. In the absence of heat shock, cortisol had no significant effect on hsp70 expression in trout hepatocytes (Fig. 5, A and B). The proteasomal inhibitors lactacystin and MG-132 significantly increased hsp70 expression in unstimulated trout hepatocytes either in the presence or absence of cortisol (Fig. 5C). The hsp70 response was significantly higher with MG-132 compared with lactacystin (Fig. 5C).

View larger version (29K): [in this window] [in a new window]   Fig. 5.   Proteasomal inhibitors increase hsp70 expression in unstimulated hepatocytes. Hepatocytes were incubated with cortisol (0, 100, or 1,000 ng/ml) and then DMSO vehicle (control; open bar), 10 µM lactacystin (crosshatched), or 50 µM MG-132 (solid bar) was added to each well and incubated for an additional 24 h. A: representative Western blot of hsp70 expression with cortisol and proteasomal inhibitors in hepatocytes; blots were probed using a trout-specific hsp70 antibody. B: bar graph showing the effect of cortisol on hsp70 expression in trout hepatocytes; values are shown as change from 100% control (0 ng/ml cortisol) and represent means + SE (n = 5 fish); there was no significant effects of cortisol (P < 0.05, 1-way ANOVA). C: bar graph showing the effect of proteasomal inhibitors on hsp70 expression either in the absence or presence of cortisol (100 or 1,000 ng/ml); values are shown as change from 100% control (0 ng/ml cortisol; open bar) and represent means + SE (n = 5 fish); 2-way ANOVA showed a significant effect of inhibitors [lactacystin (crosshatched) and MG-132 (solid bar) significantly increased hsp70 expression compared with the control (open bar); MG-132 response was significantly higher than lactacystin] (P < 0.05; 2-way ANOVA), but neither cortisol nor the interaction of cortisol with inhibitors showed any significant effect.

View larger version (24K): [in this window] [in a new window]   Fig. 6.   Proteasomal inhibitors abolish cortisol effect on hsp70 response. Hepatocytes were incubated with cortisol (0, 100, or 1,000 ng/ml), heat shocked (+15°C; 1 h), and then DMSO vehicle (control; open bar), 10 µM lactacystin (crosshatched), or 50 µM MG-132 (solid bar) was added to each well and incubated for an additional 24 h. A: representative Western blot of hsp70 expression with cortisol and proteasomal inhibitors in heat-shocked hepatocytes; blots were probed using a trout-specific hsp70 antibody. B: bar graph showing the effect of cortisol on hsp70 expression in trout hepatocytes; values are shown as change from 100% control (0 ng/ml cortisol) and represent means + SE (n = 3 fish); bars with the same letters are not statistically different (P < 0.05, 1-way ANOVA). C: bar graph showing the effect of proteasomal inhibitors on hsp70 expression either in the absence or presence of cortisol (100 or 1,000 ng/ml) in heat-shocked hepatocytes; values are shown as change from 100% control (0 ng/ml cortisol; open bar) and represent means + SE (n = 3 or 4 fish); 2-way ANOVA showed a significant cortisol effect (bars with the same letters are not significantly different) and inhibitor effect [MG-132 (solid bar) significantly increased hsp70 expression compared with the control (open bar) and lactacystin groups (crosshatched)] (P < 0.05; 2-way ANOVA), but there was no significant interaction.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our results show for the first time that chronic cortisol stimulation affects hsp70 turnover in heat-shocked trout hepatocytes. Trout hepatocytes synthesized hsp70 over a 24-h recovery period after a 1-h heat shock (Figs. 1A and 2A), and the sustained synthesis of hsp70 evident in the present study may perhaps be due to the higher intensity (+15°C) of heat shock. Despite the higher intensity of heat shock, the cell viability was not altered (4) and may be attributed to the accumulation of hsp70 (Fig. 1C), preventing proteotoxicity and allowing cells to cope with the stressor (10). The significantly lower hsp70 accumulation by cortisol (Fig. 2C) suggests that glucocorticoids may compromise hepatocyte capacity to elicit an hsp70 response in rainbow trout. The in vitro finding did concur with a recent study showing that chronic glucocorticoid treatment in vivo decreased the heat-induced hsp70 expression in rainbow trout liver (3), clearly establishing the physiological relevance of our hepatocyte system. Also, this observation was not unique to trout because other species of fish (3) and rat brain slices (2), as well as transformed cell lines (27), showed a similar glucocorticoid-mediated heat-shock response.

The obvious mechanisms for attenuation of the hsp70 response with cortisol would include either decreased synthesis and/or increased degradation of hsp70. ³⁵S-labeling studies did show that cortisol lowered heat shock-induced nascent hsp70 synthesis (Fig. 2A), arguing for a sustained attenuation of hsp70 synthesis over a 24-h period. This downregulation of hsp70 synthesis could account for the significant drop in hsp70 accumulation by cortisol in heat-shocked hepatocytes (Fig. 2C). Also, the lack of a cortisol effect on [³⁵S]hsp70 decay (Fig. 3C) clearly argues against hsp70 degradation as a possible cause for the attenuation of the heat shock response. Together these results suggest that cortisol affects hsp70 turnover by decreasing the protein synthetic capacity in heat-shocked hepatocytes. The mechanism(s) behind glucocorticoid-mediated attenuation of hsp70 synthesis is not clear; however, a recent study with mouse L929 cells implicates a GR-mediated inhibition of HSF1 as the likely cause for the attenuation of the heat-shock response (27). Also, we showed that cortisol lowered hsp90 mRNA accumulation in response to heat shock in trout hepatocytes (22), suggesting that GR-mediated hsp response may be transcriptionally regulated. If that is the case, then treatments that affect cellular GR levels may potentially modulate the heat-shock response.

In fish, including trout, cortisol treatment and/or stressors that stimulate cortisol elevation, decrease liver GR capacity in vivo (16). It is well established that chronic cortisol stimulation downregulates GR levels in vitro in mammalian system (28) and in trout hepatocytes in primary culture (present study). Also, heat shock, even in the absence of cortisol, downregulates GR in trout hepatocytes, but this decreased GR correlated with elevated hsp70 content (4). Consequently, the lower hsp70 expression in heat-shocked hepatocytes in the presence of cortisol argues for a direct ligand-mediated GR impact on hsp70 expression in the present study. As cellular glucocorticoid response is directly proportional to GR capacity (4, 28), the downregulation of GR may be a reason for the decreased hsp70 synthesis with cortisol in the present study (Fig. 2C). As the proteasome is involved in ligand-mediated GR downregulation, inhibition of the proteasome may be one possible way to increase GR levels. Indeed proteasomal inhibition increased GR expression and also GR transcriptional activity in COS-1 and HeLa cells, implicating a role for the proteasome in the GR signaling process (28). Our results are in agreement showing that the proteasome is involved in cortisol-mediated GR downregulation also in heat-shocked trout hepatocytes (Fig. 4C). However, the increase in GR expression with MG-132 and lactacystin were similar either with or without cortisol (Fig. 4C), suggesting that the GR downregulation by cortisol may not be fully accounted for by the proteasome and may also include lower GR synthesis.

Inhibiting the proteasome did result in significantly elevated hsp70 accumulation only in the absence, but not in the presence, of heat shock in trout hepatocytes (Figs. 5B and 6B). The higher hsp70 accumulation in the absence of heat shock is not surprising given the fact that inhibition of the proteasome causes accumulation/aggregation of proteins that are normally bound for degradation, in turn activating the heat-shock transcriptional machinery (5, 11, 18, 19). Indeed, the significantly higher hsp70 response with MG-132 compared with lactacystin provides further support to this argument because MG-132 is a general inhibitor of the proteasome, resulting in higher protein accumulation relative to lactacystin, which is substrate specific. If this were the case, then with heat shock we should have seen a higher hsp70 accumulation with the inhibitors, but the lack of any change may imply inhibition of the proteasome with heat shock in trout hepatocytes, favoring hsp70 accumulation (Fig. 1C). The mechanism(s) associated with heat shock-mediated proteasomal inhibition is not clear, but studies showed that heat shock locks the 20S proteasome in an inactive state and reduces ATP-dependent activation of the 26S complex (12). Also, heat shock reduced mRNA levels of proteasomal subunits including MC3 (12). We showed that GR accumulates in heat-shocked trout hepatocytes treated with the proteasomal inhibitors (Fig. 4C), clearly arguing against a complete heat shock-induced inhibition of the proteasome, instead raising the possibility that the heat-shock response may be specific to hsp70 degradation.

Inhibiting the proteasome did result in significantly elevated hsp70 accumulation in the cortisol group compared with the control group in heat-shocked trout hepatocytes (Fig. 6C). This would suggest that cortisol is mediating hsp70 breakdown via the ubiquitin-proteasome pathway in heat-shocked trout hepatocytes. However, that appears unlikely given the fact that direct measurement of [³⁵S]hsp70 decay showed no significant effect of cortisol on hsp70 breakdown in trout hepatocytes (Fig. 3C). Therefore, it appears likely that the higher hsp70 content with cortisol is due to increased protein synthesis in these cells. Furthermore, the higher hsp70 accumulation with the proteasomal inhibitors in the cortisol group correlated with a higher GR content in these cells (Figs. 4C and 6C), suggesting a role for GR signaling in the regulation of hsp70 response in trout hepatocytes. This notion finds support from a recent study showing that proteasomal inhibition not only increased GR levels but also enhanced GR transcriptional activity in mammalian cell lines (28).

The GR response to cortisol stimulation was similar with both the proteasomal inhibitors, whereas the hsp70 response was greater with MG-132 than with lactacystin (Figs. 4C and 6C), implying that perhaps other proteins, including transcriptional factors, whose breakdown is inhibited by MG-132, may be involved in the GR-mediated transcriptional regulation of hsp70 in trout hepatocytes. Taken together, these results clearly establish the importance of the GR signaling pathway in the heat-shock response in trout hepatocytes. The proteasome may be playing a key role in the GR-mediated transcriptional regulation under stressful conditions, especially those resulting in chronic cortisol stimulation. This may include proteasome-mediated repression of the ligand-dependent GR transcriptional signaling as evidenced in studies with transformed cell lines (28). Also, the observation that glucocorticoid inhibits hsp70 transcription via HSF1 (27) suggests other possible mechanisms in addition to GR downregulation that may modulate hsp transcription in stressed cells. The concentration of cortisol used in the present study mimics physiological levels seen in healthy unstressed humans [during their diurnal cycle (24)], suggesting that the heat shock response may follow a diel pattern in humans. The clinical implications are that the cellular stress-coping mechanisms, especially the role of molecular chaperones in protecting cells from proteotoxicity, may be compromised in individuals either on a daily basis (associated with diel changes in cortisol levels) and/or due to chronic cortisol overstimulation associated with stress and/or pathological states. As glucocorticoids are playing an important role in regaining homeostasis poststress (16, 21), it remains to be determined whether the decreased hsp70 response with cortisol is either adaptive and/or maladaptive.

In conclusion, our study showed that elevated cortisol levels, typically seen in animals exposed to chronic stressors, affect hsp70 turnover in heat-shocked cells. Specifically, cortisol decreased hsp70 synthesis and lowered the hsp70 accumulation evident after a heat shock in trout hepatocytes. We provide evidence for GR protein degradation, mediated by the proteasome, as a mechanism for the attenuated hsp70 response by cortisol. Our study establishes for the first time a mechanistic link between the stress-induced endocrine response and the cellular heat-shock response using a physiologically relevant cell system.