INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

STRESS OR HEAT SHOCK proteins (HSPs) are a conserved group of inducible and constitutive proteins that are thought to protect cells from stresses that typically result in protein malfolding, such as hypoxia, ischemia, reactive oxygen species, or heavy metals (28, 31). The constitutive expression of these proteins indicates that, in addition to their protective function, they also have an important role in normal cell functioning, acting as molecular chaperones (14). Considerable information exists concerning the role of HSPs in folding and protection; however, the physiological and biochemical signals that induce their synthesis remain conjectural. It is noteworthy that stressful conditions are frequently accompanied by disturbances in energy metabolism, typically assessed as changes in ATP concentrations. Therefore, alterations in energy metabolism may have the potential to affect HSP function and expression at several levels.

ATP depletion is thought to cause accumulation of denatured proteins (30) and aggregation of constitutive HSP70 that ultimately triggers the heat shock response under non-heat-shock temperatures (1, 2, 25). It has also been hypothesized that during metabolic stress (i.e., decrease in ATP), the constitutive and inducible forms of HSP70 become stably bound with proteins that are unable to fold or have become denatured. As the available HSP70 levels decline, the cells respond by increasing HSP expression through a feedback mechanism yet to be described (3, 5, 29).

Several studies document a correlation between decreases in cellular energy (i.e., ATP) and HSP activation. ATP depletion was shown to be sufficient to activate the heat shock response in HeLa cells (24) and C2C12 myogenic cells (5). Ehrlich tumor cells (11), murine cell lines (30), and rat cardiomyocytes (34) also demonstrated increased synthesis of HSP70 in response to ATP depletion arising from metabolic inhibition. In contrast, some experimental models do not exhibit a correlation between ATP depletion and HSP expression. Hypoxic rat cardiomyocytes increased HSP gene expression before any significant decrease in ATP was observed (16). Furthermore, although salicylate-induced decreases in ATP levels activated heat shock factor (HSF) binding to DNA in Drosophila salivary glands, HSP gene transcription was inhibited (35). These studies have correlated changes in ATP with changes in HSP expression. However, in tissues where ADP and AMP can be further metabolized, ATP concentration alone is not a valid indicator of bioenergetic stress. In these tissues, ATP-to-ADP ratio (ATP/ADP) is a more sensitive indicator of changes in energy metabolism (e.g., Ref. 32). Thus the effect of cellular energy status on HSP synthesis remains equivocal.

In the present study, we use nucleated red blood cells from rainbow trout to address the interaction between energy status and HSP gene expression. We previously showed that thermal stress causes an induction of HSP70 synthesis in these cells, whereas severe hypoxia does not (9). It is expected that thermal and hypoxic stress would each cause deleterious or compensatory changes in energy metabolism. We have investigated this paradox by studying, in parallel, changes in HSP expression (HSP70, HSC71, HSP30) and energy metabolism (lactate production, ATP, and ADP) in response to high temperature, severe hypoxia, and inhibitors of oxidative phosphorylation and glycolysis. Collectively, these studies suggest that HSP mRNA expression does not increase as a result of bioenergetic stress in nucleated red blood cells. We have concluded that the physiology and biochemistry of the cell/tissue type is critical when evaluating stress with HSPs.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals. Male and female adult rainbow trout, Oncorhynchus mykiss, (500-1,000 g) were obtained from Pure Springs Trout Farm in Belleville, Ontario, and were transported to Queen's University where they were maintained in dechlorinated freshwater tanks. The fish were acclimated at 10°C (±1°C) for at least 1 mo before experiments.

The trout were anesthetized in a buffered 3-aminobenzoic acid ethyl ester (MS-222, Sigma) fresh water mixture (1:10,000), and the dorsal aorta was cannulated using the method of Smith and Bell (33). After surgery, the fish recovered overnight in darkened acrylic boxes supplied with aerated flowing water at 10°C.

Inhibitors. To block aerobic ATP production, sodium azide (Az), an inhibitor of cytochrome oxidase, was added to the blood to a final concentration of 5 mM. The glycolytic inhibitor iodoacetic acid (IAA; final concentration 10 mM) was added to the blood in combination with Az. We measured lactate production and ATP and ADP concentrations using several doses of Az and IAA and determined that these concentrations were optimal.

Experimental design. Whole blood (8-14 ml) was removed from the dorsal aortic cannula of the fish and placed at 10°C in a round-bottomed flask with ~2 ml of physiological saline (in mmol/l: 124.1 NaCl, 5.1 KCl, 1.9 MgSO₄, 1.5 Na₂HPO₄, 11.9 NaHCO₃, 1.1 CaCl₂, 5.6 glucose) containing 20 mmol/l EDTA to prevent clotting. In the three experimental series that are described below, whole blood (hematocrit ~25%) was aliquoted among round-bottomed flasks, treated with the appropriate inhibitor (if applicable), and held at 10°C for 10 min. Blood was then sampled (1 or 2 ml) for either metabolites or RNA analyses, as described below. This sample represented time 0. The flasks were then placed under their respective gas and temperature conditions and rotated at ~75 rpm. A 1-ml blood sample was taken for metabolite analyses after 15 min of equilibration. Samples for both metabolites and RNA (1 ml each) were taken after 2 h of equilibration.

In the first experimental series, 2 or 3 ml of whole blood were aliquoted to each of four flasks. Two flasks were treated with Az and two flasks remained untreated. At time 0, two flasks (Az and untreated) were transferred to 25°C and two flasks (Az and untreated) were kept at 10°C. Blood was sampled from each treatment for later analysis of metabolites or mRNA after 15 min (metabolites only) and 2 h of equilibration. This experimental series was repeated under hypoxic conditions (100% nitrogen), with the exception of the Az treatment.

In the final experimental series, 2 ml of whole blood were aliquoted to each of four flasks kept at 10°C. Az was added to two flasks and one of these flasks was also treated with IAA. Three flasks (untreated, Az, Az + IAA) were placed at 10°C while the remaining flask (untreated) was placed at 25°C. Blood was sampled from each treatment for later analysis of metabolites and mRNA after the 2-h equilibration period.

RNA isolation and Northern blot analysis. Total RNA was extracted according to the acid phenol method of Chomczynski and Sacchi (8). Total RNA (10 µg) was fractionated by glyoxal-DMSO-denaturing electrophoresis on a 1.2% agarose gel and capillary transferred to a nylon membrane (Stratagene Duralon) using 20× standard saline citrate (SSC). Membranes were ultraviolet-crosslinked (Fisher UV Crosslinker) twice at the optimal setting before hybridization with probes.

HSP probes were generated using first-strand cDNA reverse transcribed from total RNA obtained from heat-shocked rainbow trout blood. The rainbow trout HSP70 probe was a 237-bp fragment between nucleotides 30 and 267 from the rainbow trout cDNA sequence pTHS70.7 (18). This probe was amplified with the forward primer 5'-CTTACCTGGGCCAGAAGGT and the reverse primer 5'-ATGGTCAGGATGGACACGT using an annealing temperature of 58°C. The rainbow trout HSC71 probe was a 240-bp fragment between nucleotides 4817 and 5252 (exons 8 and 9) from the trout HSC71 gene (36). Amplification used the forward primer 5'-TAAGATTCTGGAGAAGTGCAA and the reverse primer 5'-TCAATGGTTGGTCCAGATGAT at an annealing temperature of 56°C. The HSP30 probe was a 630-bp fragment between nucleotides 43 and 673 from chinook salmon (Oncorhynchus tschawytscha) cDNA sequence (Genbank accession no. U19370). The forward primer 5'-GAGGATTCCAGTCTTCCCTCAG and the reverse primer 5'-CAGTCCTCCTTTGAAATGGTG were used to amplify this probe at an annealing temperature of 60°C. The identity of all PCR products was confirmed by sequencing. Drosophila -tubulin was generously donated by Dr. W. Bendena (Queen's University).

The probes were labeled using [³²P]dCTP (specific activity of 1 × 10⁹ counts · min¹ · µg DNA¹) and the Ready-To-Go labeling system (Pharmacia). Membranes were prehybridized at 42°C for 2-4 h in a solution containing 25 mM K₂HPO₄, 50% formamide, 5× SSC, 5× Denhardt's reagent, and 50 µg/ml denatured salmon sperm DNA. Hybridizations were performed in prehybridization solution plus dextran sulfate (50 µM) at 42°C for 18 h with ~10⁹ counts/min denatured probe, and blots were washed at 65°C twice each in solutions of 1× SSC-0.1% SDS and 0.25× SSC-0.1% SDS. Autoradiography was performed with intensifying screens using NEN reflective film and quantified using a Molecular Dynamics computing densitometer with associated Molecular Dynamics ImageQuant (version 3.3) software. Multiple exposures were performed to ensure linearity of the film. Membranes were probed with -tubulin to control for variations in loading. The ratios of HSP70 to tubulin, HSC71 to tubulin, and HSP30 to tubulin were set at an arbitrary value of 1 in the control condition (10°C in air).

Metabolite assays. In preparation for metabolite assays, whole blood samples (1 ml) were acidified with 0.1 ml of 70% perchloric acid and centrifuged 5 min at 15,000 g. The supernatant was neutralized by addition of saturated Tris base (~0.3 ml) and 2 M KOH (~0.2 ml). Metabolites were measured spectrophotometrically at 340 nm using enzyme-coupled assays (22) modified for use on a Molecular Devices Spectramax Microplate Spectrophotometer.

Lactate was measured in samples and standards using lactate dehydrogenase (LDH). A hydrazine trap (in mM: 200 hydrazine, 500 glycine, 1 EDTA, pH 9.5) was used to drive lactate oxidation to completion. The reaction mix contained 2 mM NAD⁺ and 1 unit (U) LDH.

The ATP assay used the coupling enzymes glucose-6-phosphate dehydrogenase and hexokinase (HK). Addition of 1 U HK in the presence of excess glucose (10 mM) and NAD⁺ (2 mM) initiated the reaction.

ADP was determined by measuring the disappearance of NADH in the presence of LDH and pyruvate kinase (PK). Preincubation in the absence of PK was used to deplete the reaction mixture of pyruvate arising from the sample and from potential phosphoenolpyruvate degradation.

Statistics. A one-way ANOVA (P < 0.05) followed by Student-Newman-Keuls all pairwise comparison test (P < 0.05) was used to assess significant differences among treatment groups for metabolites, and t-tests were used to determine significant differences (P < 0.05) between 10 and 25°C. A two-way ANOVA (P < 0.05) followed by Student-Newman-Keuls all pairwise comparison test (P < 0.05) determined the significance of temperature and treatment for relative HSP band densities. A one-way ANOVA (P < 0.05) was also used to determine differences among HSP band densities in hypoxia and IAA experiments.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Metabolic stress in rainbow trout red blood cells. To define a metabolic stress in the rainbow trout red blood cell, we have characterized the energetic state of the cell under non-stress conditions. Metabolic rate was estimated as the sum of glycolytic and oxidative processes (Table 1). When oxidative phosphorylation is inhibited with Az, it is presumed that glycolysis and the existing pools of ATP are used to meet the demand of energy metabolism. Thus the increase in lactate concentration and decrease in ATP concentration due to Az treatment represents the energy that is supplied by oxidative phosphorylation under control conditions. It should be noted that these calculations assume the metabolic rate of the cell remains constant under Az conditions. If we also assume an ATP/O₂ of 6 (15), our estimated values of metabolic rate are well within range of the oxygen consumption values reported previously for rainbow trout red blood cells (10). With the use of these calculations, ~80% of the cell's metabolic rate may be attributed to oxidative phosphorylation under non-stress conditions and this contribution is not significantly affected by heat shock (Table 1). Equilibration for 2 h under control conditions at 10°C significantly decreased ATP/ADP by 26% (Fig. 1A); however, there was no change in ATP concentration (Fig. 1B).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   ATP-to-ADP (ATP/ADP; A), ATP concentrations ([ATP]; B), and lactate (Lac) production (C) in rainbow trout whole blood under control conditions (air), hypoxia, and after azide treatment. Blood was equilibrated in air, 100% nitrogen (hypoxia), and with a final concentration of 5 mM sodium azide after 120 min (n = 12) at 10 or 25°C. In A and B, control was sampled at 10°C before equilibration, and in A all values were significantly lower than the control. Lactate production is significantly greater at 25°C in all treatments. * Significant (P < 0.05) difference from air;  significant (P < 0.05) difference from the hypoxic condition; + significant (P < 0.05) differences between 10 and 25°C. All values are expressed as means + SE. rbc, Red blood cells.

Effects of heat shock and hypoxia on bioenergetics and HSP mRNA expression. A 2-h heat shock of 25°C did not significantly affect ATP concentration but resulted in a significant decrease (35%) in ATP/ADP relative to 10°C (Fig. 1A) and a significant increase (74%) in lactate production (Fig. 1C) in rainbow trout red blood cells. These metabolic alterations were accompanied by a significant increase in expression of HSP70 but not HSP30 mRNA (Figs. 2, A and B, and 3). Rainbow trout red blood cells also transcribe the constitutive member of the HSP70 family, HSC71, and this transcription was not significantly affected by heat shock (Table 2). Hypoxia alone did not induce HSP transcription in these cells, but cells that were heat shocked under severe hypoxic conditions demonstrated significant HSP70 (Fig. 2C) and HSP30 (data not shown) expression that was similar in magnitude to heat shock alone. However, both ATP and ATP/ADP were significantly lower than heat shock alone (Fig. 1, A and B).

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Quantification of heat shock protein (HSP) 70 and 30 mRNA transcription in rainbow trout whole blood under control conditions (air), after treatment with azide, and in hypoxia. In A, blood was equilibrated in air and in air with a final concentration of 5 mM sodium azide (Az) at both 10 and 25°C. Bar graphs depict relative HSP70 (A) and HSP30 (B) mRNA band density from densitometric scans of Northern blots from 10 and 5 individuals, respectively. In C, blood was equilibrated with 100% nitrogen (hypoxia) at both 10 and 25°C. C, control condition (time 0 in air). Bar graph represents relative HSP70 band density from scans of Northern blots from 4 individuals. All values were corrected for possible loading variation with Drosophila -tubulin and then expressed relative to the control condition (time 0 10°C in air), which was assigned an arbitrary value of 1. * Significant (P < 0.05) difference from control (C), + Significant difference from 10°C. All values are expressed as means + SE.

View larger version (81K): [in this window] [in a new window]   Fig. 3.   Representative autoradiograph from a Northern blot probed with HSP70 (top) and HSP30 (bottom) using RNA from rainbow trout whole blood. Several blots were probed for both HSP70 (n = 10; Fig. 2) and HSP30 (n = 5; Fig. 2) and yielded similar results. Blood was equilibrated for 2 h at 10 or 25°C under control conditions and in the presence of 5 mM Az. Blood was sampled before equilibration (10°C time 0) and after 2 h for the remainder of the treatments. Other details are described in Experimental design.

Inhibition of oxidative phosphorylation and glycolysis does not induce HSP gene transcription. To investigate the relationship between energy status and HSP transcription we measured HSP gene transcription after treatment of rainbow trout red blood cells with metabolic inhibitors. Inhibition of oxidative phosphorylation with Az decreased ATP by 56% (Fig. 1B) and increased lactate production by 62% relative to 10°C in air (Fig. 1C) but did not induce HSP gene transcription (Figs. 2, A and B, and 3). Az also resulted in a greater decrease in ATP than that observed under normoxic or hypoxic conditions at both temperatures after 15 (data not shown) and 120 min (Fig. 1B). Az did not significantly induce expression of HSP70 or HSP30 mRNA expression after heat shock (Figs. 2, A and B, and 3) but resulted in a significantly greater decrease in ATP (Fig. 1B) and a significantly greater increase in lactate production (Fig. 1C) relative to heat shock.

Inhibition of oxidative phosphorylation with Az and glycolysis with IAA resulted in a >75% decrease in ATP (Fig. 4B). Lactate production was prevented (Fig. 4C), attesting to the efficacy of the IAA treatment. Despite these bioenergetic alterations, there was no induction of either HSP70 (Fig. 5) or HSP30 (data not shown) gene transcription in these cells.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   ATP/ADP (A), ATP concentrations (B), and lactate production (C) from rainbow trout whole blood demonstrating the effects of treatment with Az and a combination of Az and iodoacetic acid (IAA). Blood was equilibrated at 10°C for 0 min (time 0) or 2 h at 1) 10°C, 2) 10°C with a final concentration of 5 mM Az (10°C + Az), 3) 10°C with both Az (final concentration 5 mM) and IAA (final concentration 10 mM), or 4) 25°C. These experiments were performed on 6 individuals. Data from these graphs (10°C, 25°C, and 10°C + Az) were also pooled and presented in Fig. 1. * Significant (P < 0.05) difference from time 0 (A, B) or 10°C (C); + significant (P < 0.05) difference from 10°C after 2 h of equilibration. All values are means + SE.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Quantification of HSP70 mRNA transcription in rainbow trout whole blood demonstrating effects of Az treatment and a combination of Az and IAA treatment. Blood was equilibrated at 10°C for 0 min (time 0) or 2 h at 1) 10°C, 2) 10°C with a final concentration of 5 mM Az (10°C + Az), 3) 10°C with both Az (final concentration 5 mM) and IAA (final concentration 10 mM), and 4) 25°C. Bar graph depicts relative HSP70 mRNA band density from densitometric scans of Northern blots from 6 individuals. All values were corrected for possible loading variation with Drosophila -tubulin and then expressed relative to the control condition (10°C) that was assigned an arbitrary value of 1. Heat shock condition (25°C) is significantly different from all other treatments (P < 0.05). All values are expressed as means + SE.

The transcription of the constitutive form of HSP70, HSC71, was not significantly affected by severe hypoxia, inhibitors, or any combination of these treatments (Table 2).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Metabolic stress, characterized by a depletion of cellular ATP, has been proposed to be an important component of the trigger that induces stress protein synthesis. A reduction in ATP is thought to cause the HSF to bind to the heat shock element in the promoter region of the heat shock gene, thereby inducing the heat shock response (3, 28, 29). We have shown that heat shock alone has no significant effect on ATP concentrations but induces increased gene transcription of the major stress protein, HSP70, in rainbow trout red blood cells (Figs. 2 and 3). Although HSP30 mRNA also increased after heat shock, the increase was not statistically significant, which may be attributed to the high variability associated with these data. Despite the maintenance of ATP levels after heat shock, however, there is a significant decline in ATP/ADP (Fig. 1A). This decrease is indicative of an increase in ADP that accelerates energy turnover and equilibrates ATP production with ATP consumption. Therefore, a decrease in ATP/ADP (increase in ADP) signals an energetic deficiency and results in a regulatory response. Any alteration in the adenylate ratio indicates that metabolism is changing, regardless of ATP concentration ([ATP]). Thus, in our study, heat shock alone probably does not cause a metabolic stress in the cell, but accelerates the rate of cellular respiration to maintain constant ATP levels. It is therefore important to examine both [ATP] and ATP/ADP to obtain a more veritable depiction of the energetic state of the cell. In terms of metabolic stress and HSP induction, however, analysis of ATP levels alone is probably sufficient.

Hypoxia causes a greater metabolic disturbance than heat shock alone, as indicated by the decrease in ATP/ADP (Fig. 1A), the depletion of ATP (Fig. 1B), and the increase in lactate production (Fig. 1C). Hypoxia has also been shown to induce stress proteins in many cells and tissues (13, 16, 17, 27), however, the results from the present study demonstrate that severe hypoxia has no affect on transcription of either HSP70 or HSP30 (Fig. 2C) in rainbow trout red blood cells. Our previous study also demonstrated that this severe hypoxic condition did not result in HSP70 protein synthesis in these cells (9). These results may be somewhat surprising considering that, unlike their mammalian counterparts, the nucleated red blood cells of lower vertebrates rely heavily on oxidative metabolism for energy (72-80% in the present study).

Changes in ATP have the potential to affect several aspects of HSP function and expression. In fact, stress proteins have an important role in energy-dependent processes, such as protein translocation into cellular organelles (4). Moreover, the binding and release of denatured proteins and short peptides to HSP70 is dependent on the hydrolysis of ATP (26). ATP is thus considered essential for proper HSP70 functioning (29). The limited information that exists on the role of ATP in the function of the small HSPs (i.e., HSP30) suggests that these HSPs do not require ATP for their function (7). As was the case with severe hypoxia, metabolic inhibition with Az and a combination of Az and IAA resulted in a significant decline in [ATP] and ATP/ADP (Figs. 1 and 4) without any significant effect on HSP70 or HSP30 mRNA (Figs. 2 and 5). This may not be unexpected in the case of HSP30 synthesis, because the few data that are available on the small HSPs indicate that ATP depletion may actually render the stress protein inactive (21).

In addition to the role of ATP in HSP function, HSP70 expression has been shown to correlate with ATP depletion in several mammalian cells and this energetic stress is thought to trigger HSP synthesis (3, 16, 30, 34). This trigger relies on the assumption that the release of HSP70 from proteins is effectively blocked by depletion of intracellular ATP and the subsequent reduction in available HSP70 induces HSP expression (3). In the present study, however, it is unlikely that ATP would ever limit HSP70 release from proteins because the Michaelis-Menten constant for ATP [1-2 µM (26)] is several orders of magnitude lower than the lowest ATP level observed (Fig. 1B). Moreover, other studies, including the present one, demonstrate no correlation between reduced ATP levels and increased HSP70 gene transcription (16, 17, 35). There are examples where reductions in ATP will activate HSF binding but will not induce HSP transcription (5, 27, 35), which may be related to the particular HSF involved or an energy requirement at several steps in the activation of transcription (27). Although we cannot comment on HSF activation in our study, the absence of HSP transcription during severe hypoxia or metabolic inhibition observed in this study suggests that ATP-independent activation pathways of the stress response must exist.

The energetic cost of mounting a heat shock response is largely unknown. Transcription and translation are energetically expensive processes, and, in fish, protein synthesis is estimated to cost 40% of the whole animal oxygen consumption (23). We have shown that rainbow trout red blood cells will not increase HSP transcription after metabolic inhibition alone (low ATP), indicating that a decrease in ATP is not part of the HSP induction mechanism in these cells. The lack of a significant heat shock response in metabolically inhibited cells (Fig. 2, A and B) suggests that under these conditions of extreme energetic stress, low ATP levels may actually limit HSP expression. Under moderate but significant decreases in ATP (i.e., heat shock and hypoxia), however, the heat shock response does not appear to be energetically limited because rainbow trout red blood cells are still capable of significantly increasing HSP70 (Fig. 2C) and HSP30 (data not shown) gene expression.

Most mammalian tissues are susceptible to the cellular damage caused by energetic stress, such as hypoxia and/or ischemia, and will respond by synthesizing HSPs (6, 27, 28). It is clear from these studies that energy stress of the magnitude observed in this study is insufficient to elicit a heat shock response in fish red blood cells. Rainbow trout red blood cells, however, may commonly be exposed to reduced oxygen levels in the circulation, particularly postexercise, and thus may be inherently hypoxia tolerant (9) as was also suggested by Ferguson et al. (10). Unlike many mammalian cells, trout red blood cells may therefore be able to cope with reduced ATP levels without requiring the protection of HSPs because hypoxia is not regarded as a potential stress. Another tissue of the rainbow trout, the myocardium, also failed to increase HSP70 synthesis after short-term hypoxic exposure (12). In addition, anoxia-tolerant turtle (Chrysemys picta bellii) hepatocytes do not synthesize stress proteins in response to reduced oxygen levels (19, 20), further suggesting that the physiology of the organism and/or tissue may be critical when assessing stress with HSPs.

In conclusion, we have shown that a reduction in cellular energy status will not induce stress protein gene transcription in rainbow trout red blood cells. This suggests there may be multiple activation pathways for the stress response, some of which may be ATP independent. Indeed, during extreme energetic stress, such that occurs during a combined heat shock/metabolic inhibition, HSP mRNA expression actually appears to be energetically limited. The relationship between ATP and HSP expression observed in other studies may be related to cell-type specific differences in gene regulation. Nucleated red blood cells, however, appear to possess strategies that allow them to cope with reduced levels of ATP without HSP expression, thereby demonstrating that the physiology of the tissue and organism must be considered when evaluating stress protein gene transcription and synthesis. Moreover, the absence of increased HSP gene transcription in response to ATP depletion indicates that stress proteins may not be valid indicators of metabolic stress.

Perspectives