Aging alters cellular responses to both heat and oxidative stress. Thiol-mediated metabolism of reactive oxygen species (ROS) is believed to be important in aging. To begin to determine the role of thiols in aging and heat stress, we depleted liver glutathione (GSH) by administering l-buthionine sulfoximine (BSO) in young (6 mo) and old (24 mo) Fisher 344 rats before heat stress. Animals were given BSO (4 mmol/kg ip) or saline (1 ml ip) 2 h before heat stress and subsequently heated to a core temperature of 41°C over a 90-min period. Liver tissue was collected before and 0, 30, and 60 min after heat stress. BSO inhibited glutamate cysteine ligase (GCL, the rate-limiting enzyme in GSH synthesis) catalytic activity and resulted in a decline in liver GSH and GSSG that was more pronounced in young compared with old animals. Catalase activity did not change between groups until 60 min after heat stress in young BSO-treated rats. Young animals experienced a substantial and persistent reduction in Cu,Zn-SOD activity with BSO treatment. Mn-SOD activity increased with BSO but declined after heat stress. The differences in thiol depletion observed between young and old animals with BSO treatment may be indicative of age-related differences in GSH compartmentalization that could have an impact on maintenance of redox homeostasis and antioxidant balance immediately after a physiologically relevant stress. The significant changes in antioxidant enzyme activity after GSH depletion suggest that thiol status can influence the regulation of other antioxidant enzymes.
- reactive oxygen species
aging has long been known to reduce an organism’s capacity for coping with repeated physiological challenges, including heat stress (14, 15, 23, 55, 56). The free radical theory of aging has gained popularity as the importance of the redox state of the cell and the age-related accumulation of oxidatively damaged and modified proteins become more well defined (2, 5). Oxidative modifications to cellular components can enhance, diminish, or alter their biological activity and reduce cellular defense and repair mechanisms (2). The glutathione (GSH) system is thought to be one of the major cellular redox buffering systems in mammals. Hydrogen peroxide (H2O2) formed by the dismutation of superoxide (O2−·) is catabolized at the expense of reducing equivalents from GSH by glutathione peroxidase (GPx) to form glutathione disulfide (GSSG) or mixed protein disulfides. It is thought that GSH depletion occurs with aging (17), although the effects of old age on rodent liver GSH appear to be strain dependent. Although liver GSH depletion has been noted in aged Wistar rats (1, 9, 25, 35, 41, 42), responses in aged Fischer 344 (F344) rats are less clear. An age-related reduction in protein, mRNA, and activity of glutamate cysteine ligase (GCL), the rate-limiting enzyme in GSH synthesis, was associated with a 37% reduction in liver GSH in aged F344 rats (25). It also has been reported that aging has little effect on liver GSH content (24, 32) and GPx activity (20), whereas GSSG content is elevated (24), reflecting an age-associated increase in oxidative stress in F344 rats. Increased hepatic GSSG is accompanied by elevated GSH, resulting in little change in the GSH/GSSG ratio. Although the data are conflicting, the general pattern of responses appears to suggest a maintenance of redox balance (based on little to no change in GSH/GSSG ratios), albeit with a slight shift away from the homeostasis observed in young animals.
Cellular GSH can be depleted pharmacologically by administration of l-buthionine sulfoximine (BSO). Depletion is achieved by a combination of selective transition state inhibition of GCL and cellular export of GSH that continues even in the absence of significant GSH synthesis (37). The highly specific inhibition of GSH synthesis and lack of toxicity of acute doses of BSO make it a useful tool for studying the consequences of intracellular GSH depletion and, in the absence of other drug effects, the effects of reactive oxygen species (ROS) produced by normal metabolism (27). BSO has been used in a variety of animal models, consistently reducing intracellular GSH by 30–80% within 2–4 h of injection, and its effects persist for up to 24 h (12, 19, 21, 22, 43, 47, 48, 50, 52, 54). However, apart from observations of GSH reduction, there has been no documentation that BSO administration inhibits GCL in vivo.
Heat stress is known to produce ROS, probably due to increased flux of oxygen through the mitochondrial electron transport chain (51). Endogenous antioxidant enzymes such as catalase, copper-zinc-containing superoxide dismutase (Cu,Zn-SOD), and Mn-SOD synergize with the GSH redox system to provide a comprehensive defense against the damaging effects of oxidative stress. The homeostatic balance between antioxidant defense mechanisms appears to be critical in maintaining normal cellular function (35, 36, 41). Moreover, an observed increase in Mn-SOD activity with age suggests the mitochondria play a significant role in the age-associated increase in ROS production (30). Whereas some investigators suggest that overexpression of Mn-SOD and Cu,Zn-SOD can extend life span in Drosophila by increasing the rate of conversion of O2−· to H2O2 (49), other reports indicate that overexpression of SOD enzymes is only effective when they are expressed in combination with catalase, allowing for more effective removal of potentially harmful H2O2 produced during the SOD-catalyzed dismutation of O2−· (34).
The effect of aging on thiol metabolism during and after a physiologically relevant challenge and the contribution from other antioxidant enzyme systems have not been well defined. Although the effects of aging on heat-induced thermotolerance are well characterized, the impact of old age on maintenance of redox homeostasis in vivo during and after a single heat challenge remains to be characterized. Therefore, the purpose of this study was to determine whether GSH plays a role in regulating catalase, Cu,Zn-SOD, and Mn-SOD activity immediately after a physiologically relevant sublethal heat stress in conscious aged animals.
Young (6 mo) and old (24 mo) male F344 rats (n = 3 per age and treatment group at each time point) were obtained from the National Institute on Aging. Animals were housed in The University of Iowa Animal Care Facility, and all experimental procedures conformed to institutional animal care guidelines. All rats received food (standard rat chow) and water ad libitum.
Heat stress, drug treatments, and tissue preparation.
Animals received an intraperitoneal injection of 4 mmol/kg BSO (Sigma Aldrich, St. Louis, MO) or 1 ml of saline (control) 2 h before a heating protocol was commenced. This 2-h duration was determined to be the optimal time frame for GSH depletion via BSO in a preliminary study (data not shown). Before heating, a colonic thermistor was placed ∼8 cm beyond the anal sphincter for measurement of core body temperature (Tc). Conscious, unrestrained animals were heated to a Tc of 41°C over a 60-min period at a rate of ∼1°C per 15-min period with a heat lamp. The rate of heating was controlled by raising or lowering the lamp as necessary. Tc was then clamped at 41°C for a duration of 30 min. Control animals from each treatment were euthanized with an overdose of pentobarbital sodium (80 mg/kg ip) without heating 2 h after injection of BSO. Animals were euthanized at three time points after heating: immediately after heat stress (0 min) and 30 and 60 min after heat stress. Livers were removed and rinsed in ice-cold phosphate-buffered saline and trimmed of any adherent tissues. Liver samples were immediately frozen in liquid nitrogen for subsequent analysis.
Spectrophotometric analysis of GSH and GSSG.
Liver samples were stored in liquid nitrogen and did not undergo any freeze-thaw cycles until thiol analysis was completed. Protein was measured using the method of Lowry et al. (26). GSH and GSSG were measured according to the method of Griffith (13).
Protein was measured using the Bradford method (Bio-Rad). Activity of GCL was determined by HPLC, using a modification of the method originally developed by Nardi et al. (33, 38). Briefly, a known amount of protein was added to a reaction buffer containing 0.1 M Tris, 20 mM MgCl2, 0.75 mM l-glutamic acid, 6 mM ATP, 50 mM KCl, 6 mM DTT, and 0.75 mM cysteine. A sample was removed immediately and added to 0.5 mM ThioGlo 3 (Covalent, Woburn, MA) in acetonitrile for HPLC analysis. The reaction was incubated at 37°C. A sample was removed every 5 min up to 20 min and added to 0.5 mM ThioGlo 3 for HPLC analysis. After HPLC analysis, a linear regression curve was fitted to the five time points for each sample, and the rate of GCL appearance was calculated from the slope of this curve. GCL activity is presented in units of femtomoles of γ-glutamyl cysteine produced per milligram of protein per minute.
Antioxidant enzyme activity.
Catalase activity of cell lysates was measured spectrophotometrically using a modification of the method originally described by Beers and Sizer (3, 45). SOD activity was determined spectrophotometrically using the method of Spitz and Oberley (46), and Cu,Zn-SOD activity was distinguished from Mn-SOD activity by inhibition of Cu,Zn-SOD activity with cyanide.
Data were analyzed with a three-way ANOVA. Pairwise contrasts were performed to determine differences among age groups, drug treatment, and changes from control. Bonferroni adjusted P values were calculated for each variable to control for multiple comparisons. A P value of <0.05 was taken as significant for the main effects and each pairwise comparison. Data are presented as means ± SE.
There were no statistically significant age-related or heat stress-induced alterations in GCL activity (Fig. 1). As expected, treatment with BSO caused a significant (P < 0.05) decline in GCL activity. There was no difference between young and old animals in the effectiveness of BSO, and there was no difference between control and postheat values, indicating that heat stress had no effect on the actions of BSO.
Old saline-treated control animals had significantly less liver GSH than their young counterparts. Heat stress appeared to increase liver GSH such that there were no differences between young and old saline-treated animals after heat stress (Fig. 2A). The decrease in GCL activity with BSO treatment appeared to affect liver GSH in young more than old animals. Liver GSH was significantly reduced in young control BSO-treated animals compared with their saline-treated counterparts. This significant reduction in liver GSH in young BSO-treated animals persisted and appeared to be exacerbated in the 60 min following heat stress. By 60 min after the conclusion of heat stress, liver GSH was significantly reduced from control within the young BSO-treated group. Old BSO-treated animals also experienced a decline in liver GSH compared with the saline-treated old group, but this decline was not as severe as that observed in young BSO-treated animals. In contrast to the young BSO-treated animals, by 60 min after heat stress, the liver GSH of old BSO-treated animals appeared to be recovering, and significantly more liver GSH was retained in these animals than in young BSO-treated rats.
Aging alone was associated with a trend for a reduction in liver GCL activity in the control condition (Fig. 1), and although this reduction did not reach statistical significance, it may have been physiologically significant enough to explain the accompanying significantly lower liver GSH observed in old saline-treated control animals (Fig. 2A). Surprisingly, there were no age-related or heat stress-associated changes in liver GSSG until 60 min after heat stress, when old saline-treated animals had significantly less liver GSSG than young saline-treated animals (Fig. 2B). With the loss of liver GSH, there were slight reductions in liver GSSG in both young and old BSO-treated animals, but none of these changes were significant.
Antioxidant enzyme balance.
Aging and heat stress had little effect on catalase activity in saline-treated animals (Fig. 3). There was no change in catalase activity in old BSO-treated animals. Young BSO-treated animals also showed little change in catalase activity until 60 min after heat stress, when there was a significant increase in catalase activity that may be associated with the severe depletion of liver GSH (Fig. 2) by this time in this treatment group.
Aging alone resulted in a significant decline in Cu,Zn-SOD activity (Fig. 4). After heat stress, old saline-treated animals appeared to have a transient increase in Cu,Zn-SOD activity; however, Cu,Zn-SOD activity returned to the significantly lower levels observed in the control group by 60 min after heat stress. After BSO treatment, there was a striking, significant reduction in Cu,Zn-SOD activity in young animals both before and after heat stress. This loss of activity may be due initially to a slight nonsignificant decline in immunoreactive Cu,Zn-SOD protein in young BSO-treated animals before and immediately after heat stress (data not shown). Cu,Zn-SOD activity was resistant to the effects of BSO in old animals. Heat stress appeared to cause a transient, slight, nonsignificant increase in Cu,Zn-SOD activity in old BSO-treated animals; however, Cu,Zn-SOD activity returned to significantly lower levels than saline-treated old counterparts by 60 min after heat stress.
Mn-SOD activity was not different between saline-treated age groups and remained unaffected by heat stress in this treatment group (Fig. 5). Treatment with BSO significantly increased Mn-SOD activity in both young and old animals. Although Mn-SOD activity remained significantly elevated after heat stress, there was a progressive decline in Mn-SOD activity after hyperthermic challenge. Old BSO-treated animals continued to display a significant elevation in Mn-SOD activity by 60 min after heat stress, whereas young BSO-treated animals had returned to levels statistically similar to those of saline-treated animals. The decline in Mn-SOD activity after heat stress in BSO-treated animals was similar in both age groups; however, this decline was not statistically different from BSO-treated control values for both age groups.
Previous work from this laboratory has shown that heat stress causes increased oxidative stress and damage in the liver of young and old animals (14, 15, 55, 56). Heat-induced uncoupling of the mitochondrial electron transport chain proteins may contribute to increased free radical production during and after heat stress by increasing the leakiness of the electron transport chain (51). In the present study we have shown, by perturbation of the GSH antioxidant system with BSO, that GSH is indeed involved in oxidant removal during and after a physiological challenge, as evidenced by the progressive decline in liver GSH in young and old animals after heat stress. Interestingly, young animals experienced a far greater reduction in liver GSH after heat stress than did old animals, whereas old animals appeared to begin to recover their hepatic GSH levels by 60 min after heat stress. Both responses may be due to age-associated differences in compartmentalization of GSH.
Mitochondria are believed to be a major source of oxidants that damage cellular proteins and nucleic acids (44), and it has been shown that mitochondrial GSH is elevated in aged animals, perhaps due to increased oxidant production associated with aging (4). BSO is known to specifically deplete cytosolic GSH, whereas mitochondrial GSH remains relatively unperturbed, establishing two separate and distinct compartments of GSH localization (39). Furthermore, in vitro work has shown a differential depletion of mitochondrial GSH by diethylmaleate during hyperthermia in Chinese hamster ovary cells. It was determined that mitochondrial GSH compartmentalization was the limiting factor in cell survival after treatment (11).
Although it has been known for some time that BSO inhibits GCL in vitro (18, 28), we believe this is the first in vivo demonstration of GCL inhibition by BSO. Rather than simply showing a reduction in tissue GSH after BSO administration, as has been previously demonstrated (12, 19, 21, 22, 43, 47, 48, 50, 52, 54), we have demonstrated that a potent and severe inhibition of GCL activity occurs in the liver of BSO-treated animals. Given the lack of difference in GCL activity between age groups and with heat stress, it appears that age-related alterations in GCL activity or sensitivity to BSO are not responsible for the different responses of young and old animals to BSO treatment.
We have previously demonstrated that stark differences exist between young and old animals in terms of susceptibility to hyperthermia-induced oxidative damage at time points further removed from a second heat stress than those presented in this report (14, 15, 23, 55, 56). In the current study, we have observed that old animals have a remarkable ability to maintain cellular homeostasis immediately after a single heat stress. Young animals appeared to be more sensitive to drug treatment, which may be indicative of age-related alterations in thiol metabolism. These differences may be attributed to altered thiol compartmentalization (4), substrate availability (25), or GCL activity (10, 25). These data, combined with previous reports of severe hepatocellular injury in old animals after a second heat stress (14, 15, 55, 56), suggest that old animals may exhaust cellular defense mechanisms during an initial environmental insult, leaving little room for defense against subsequent physiological challenges.
Similar to other studies, we found no age-related alteration in catalase activity (7) in saline-treated control animals, nor did heat stress alter catalase expression or activity in young or old saline-treated animals. Given the observation that increased catalase has been shown to potentiate heat stress-induced cell death via scavenging of H2O2, thereby attenuating the H2O2-mediated activation of heat shock factor 1 (HSF1) and subsequent heat shock protein 70 induction (40), it is possible that H2O2 is a major by-product of heat stress in our model. Increased catalase activity may not be observed immediately after heat stress to allow unimpeded HSF1 activation and subsequent development of thermotolerance. It must be noted that the use of cell lysates in measuring catalase activity is, in essence, a measure of the H2O2-removing capacity of that cell lysate. Although we saw little change in catalase activity as measured spectrophotometrically, subtle changes in catalase activity cannot be excluded when these data are considered with the reduction in liver GSH.
It has been proposed that reductions in hepatic GSH are counterbalanced by increases in hepatic SOD activity in aged Wistar rats (35, 41). Although it is tenable to postulate that age-related changes in one enzyme are compensated for by alterations within another, an increase in SOD activity to counterbalance a reduction in GSH would appear to further overwhelm the system with H2O2, perhaps by increasing the dismutation of O2−·. Maintaining a balance between enzymes and proteins responsible for each step of removal of ROS would seem to be a prerequisite for maintaining redox homeostasis. Similar to other work (25), we found an age-related decrease in hepatic GSH. We had expected an increased reliance on catalase activity after treatment with BSO to counterbalance the reduction in cellular GSH and keep H2O2 levels in check (53). This outcome was not observed and may play a role in the reduction in Cu,Zn-SOD activity noted in the present study, because Cu,Zn-SOD is sensitive to H2O2-induced inactivation. Treatment with BSO was associated with a significant, sustained decrease in Cu,Zn-SOD activity in young animals. Early work showed that Cu,Zn-SOD is irreversibly inactivated by H2O2 (16). These data were collected in an in vitro system at a nonphysiological pH of 9–10, which raises questions concerning the relevance of this finding when applied to an in vivo model. However, other investigators have found a decrease in Cu,Zn-SOD activity in vivo in response to oxidative stress in iron-loaded rats (6). In combination, these data suggest that H2O2 may play an important role in vivo in modulating Cu,Zn-SOD activity.
In the present study, young BSO-treated animals experienced a greater decline in liver GSH than their old BSO-treated counterparts. Although this decline was not significantly different from old BSO-treated animals, there was a greater difference between young saline- and BSO-treated animals than between old saline- and BSO-treated animals. Similarly, young BSO-treated animals had a significant reduction in Cu,Zn-SOD activity that reflected the changes occurring in liver GSH (i.e., as liver GSH continued to decrease, so did Cu,Zn-SOD activity). Although the current data support the observation of Brown et al. (6) that Cu,Zn-SOD is not likely to be inactivated by H2O2 alone, these results do suggest that H2O2 plays a major role in inactivating Cu,Zn-SOD in vivo. There were no age-associated differences in Cu,Zn-SOD activity until 60 min after heat stress, when old saline-treated animals had significantly less Cu,Zn-SOD activity than young saline-treated animals. This observation, coupled with reductions in liver GSH and no change in catalase activity in old animals, suggests an additive effect and synergy between the antioxidant systems in the liver and provides further support for the role of H2O2 in inhibiting Cu,Zn-SOD in vivo.
In the presence of thiol-reducing chemicals such as DTT and β-mercaptoethanol, proteins of the electron transport chain are more resistant to thermal denaturation (51). Treatment with BSO has the opposite effect, which is manifested as a reduction in thiol-reducing equivalents in the form of GSH. It is likely that mitochondrial pools of GSH were minimally affected by BSO treatment (29, 39). In the presence of BSO, increased Mn-SOD presents a conundrum. Normally, H2O2 produced by the Mn-SOD-catalyzed dismutation of O2−· can diffuse down a concentration gradient into the cytosol, where it can be removed by cytosolic GSH (8). Treatment with BSO may alter cytosolic H2O2, reducing the gradient for H2O2 diffusion from the mitochondria. An imbalance in mitochondrial H2O2 and O2−· would lend itself to accumulation of O2−· and necessitate an increase in Mn-SOD activity, which may be the case in the present study. Support for this hypothesis is evident in the slightly greater reduction in GSH in young BSO-treated animals, coupled with a slightly greater increase in Mn-SOD activity. It has been hypothesized that decreasing mitochondrial O2−· may alleviate age-related oxidative damage (31), and removal of H2O2 can be accomplished by mitochondrial GSH (29). Indeed, overexpression of Mn-SOD in adult Drosophila produced an increased life span that was attributed to a more efficient dismutation of O2−· to H2O2, thereby reducing O2−· levels (49).
In summary, thiol metabolism appears to be closely linked to antioxidant enzymes responsible for removing O2−· and H2O2 from the cell, demonstrating the synergistic balance between these enzymes and thiols in maintaining cellular redox homeostasis. The significant change in antioxidant enzyme activity levels after perturbation of cellular thiols indicates that thiol status is an important factor in maintaining antioxidant balance and overall redox homeostasis. However, it appears that this balance is precarious. The large fluctuations in antioxidant enzyme activity observed in this in vivo study strongly suggest aged animals are characterized by a lack of “fine-tuning” of antioxidant enzyme activity, especially after thiol manipulations. Altered cellular localization of thiols may contribute to this dysregulation. Surprisingly, young animals appeared to be more sensitive to the addition of the GCL inhibitor BSO, likely because of the more extensive depletion of GSH (presumably cytosolic) observed in young BSO-treated animals. Further work to determine the impact of age-associated differences in cellular localization of antioxidant enzymes and proteins will be necessary to elucidate the effectiveness of interventions that influence antioxidant status.
This work was supported by National Institutes of Health Grants R01-AG-12350 (to K. C. Kregel), R01-HL-51469 (to D. R. Spitz), R01-CA-100045 (to D. R. Spitz), and P01-CA-66081 (to D. R. Spitz).
We acknowledge the administrative assistance of Joan Seye.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society