Altered accumulation of hepatic mitochondrial antioxidant proteins with age and environmental heat stress

Abstract

Aging impairs overall physiological function, particularly the response to environmental stressors. Repeated heat stress elevates reactive oxygen species and macromolecular damage in the livers of aged animals, likely due to mitochondrial dysfunction. The goal of this investigation was to determine potential mechanisms for mitochondrial dysfunction after heat stress by evaluating key redox-sensitive and antioxidant proteins (Sirt-3, MnSOD, Trx-2, and Ref-1). We hypothesized that heat stress would result in greater mitochondrial abundance of these proteins, but that aging would attenuate this response. For this purpose, young (6 mo) and old (24 mo) Fisher 344 rats were exposed to heat stress on two consecutive days. During each heating trial, colonic temperature was elevated to 41°C during the first 60 min, and then clamped at this temperature for 30 min. Nonheated animals served as controls. At 2 and 24 h after the second heat stress, hepatic mitochondria were isolated from each animal, and then immunoblotted for Sirt-3, acetylated lysine residues (Ac-K), MnSOD, Trx-2, and Ref-1. Aging increased Sirt-3 and lowered Ac-K. In response to heat stress, Sirt-3, Ac-K, MnSOD, and Ref-1 increased in mitochondrial fractions in both young and old animals. At 2 h after the second heat stress, mitochondrial Trx-2 declined in old, but not in young animals. Our results suggest that some components of the response to heat stress are preserved with aging. However, the decline in Trx-2 represents a potential mechanism for age-related mitochondrial damage and dysfunction after heat stress.

NEW & NOTEWORTHY Our results suggest heat stress-induced mitochondrial translocation of Sirt-3, MnSOD, and Ref-1 in young and old animals. Aged rats experienced a decline in Trx-2 after heat stress, suggesting a potential mechanism for age-related mitochondrial dysfunction.

INTRODUCTION

July 2023 was the warmest month ever recorded, with records going back to 1880 (1). In addition, the global mean temperature has been on an upward trajectory since the 1970s (2). Heat causes the greatest number of fatalities of any weather event (3) and elderly individuals are at the greatest risk for morbidity and mortality during heat waves (4–6). Therefore, the continuance of higher temperatures worldwide necessitates investigating physiological and cellular responses to hyperthermia to improve outcomes after periods of extreme heat. Mortality from heat waves results primarily from cardiovascular collapse (7), yet hyperthermia also elicits damage to other organs that could adversely affect health after heat waves in survivors. For example, the liver is a prime target of heat-induced injury in humans (8, 9), and our laboratory has demonstrated that the aged rodent liver experiences greater inflammation and oxidative injury after a repeated heat stress protocol (10).

At the subcellular level, liver mitochondria are particularly sensitive to hyperthermia; aged animals demonstrate augmented ultrastructural damage and compromised ATP production after environmental heat stress (11, 12). Since mitochondria regulate several essential processes in the cell including energy transduction, calcium metabolism, steroid hormone biosynthesis, and apoptosis, proper mitochondrial function is essential to the survival of an organism. In response to conditions that challenge organismal and cellular homeostasis, mitochondria exhibit a unique stress response, which includes the import of proteins such as heat-shock proteins (Hsp) 10, and 60, as well as mitochondrial Hsp70 [mtHsp70; (13, 14)]. These proteins maintain the integrity of other proteins in the mitochondria, and assist with protein import; thus they are crucial for overall mitochondrial homeostasis. Our laboratory has previously investigated the mitochondrial Hsp response in the contexts of aging and environmental heat stress, demonstrating that aging was associated with blunted mitochondrial accumulation of Hsp10, Hsp60, and mtHsp70 after exposure to hyperthermia (11).

In the liver, aging is also associated with increased mitochondrial superoxide production in the control condition (11) and augmented oxidative stress and cellular injury in response to hyperthermia (10). These changes in redox state and mitochondrial homeostasis with aging and hyperthermia prompted us to investigate the regulation of redox-sensitive and antioxidant proteins in the mitochondria in this study. Specifically, we were interested in characterizing the abundance of the mitochondrial deacetylase enzyme, Sirtuin-3 (Sirt-3) because it is a redox-sensitive, stress-inducible protein (15). In HEK293 cells and mouse embryonic fibroblasts, Sirt-3 protects against mitochondrial damage via deacetylation of isocitrate dehydrogenase-2 (IDH-2) and maintenance of the reducing agent, NADPH (16). Moreover, in the liver, Sirt-3 deacetylates manganese superoxide dismutase (MnSOD) and protects against radiation-induced apoptosis (17). While the activity of Sirt-3 and its expression in the liver is affected by both aging and caloric restriction (18), less is known about its regulation in response to physiological stress in vivo. Since Sirt-3 is the primary mitochondrial deacetylase enzyme (19), we also characterized acetylated lysine residues of mitochondrial proteins (Ac-K) as a potential read-out for Sirt-3 activity - these measures are expected to be inversely proportional. Acetylated lysine residues accumulate in the liver after a high-fat diet and ethanol intoxication (20, 21), yet it is unclear whether they are altered by hyperthermia.

In addition to Sirt-3 and Ac-K, we characterized the abundance of three key mitochondrial antioxidant proteins: MnSOD, thioredoxin-2 (Trx-2), and apurinic/apyrimidinic endonuclease-1/redox factor-1 (APE-1/Ref-1). These proteins are crucial for survival as demonstrated by either neonatal or embryonic lethality of homozygous knockout mice (22–24). MnSOD scavenges mitochondrial superoxide production, and its activity increases in the liver after heat stress in vivo (25); however, whether the increase in activity is due to elevated mitochondrial protein in this model remains unknown. The importance of MnSOD in the cellular response to hyperthermia is also emphasized by the finding that prostate and breast cancer cells overexpressing MnSOD exhibit greater survival to acute hyperthermic exposures (26, 27).

Through its thiol-reducing activity (28), Trx-2 protects against oxidative stress; haploinsufficiency of Trx-2 results in greater amounts of oxidized proteins, lipids, and DNA in the liver (29). In addition, Trx-2 suppresses mitochondrial reactive oxygen species production, and its ablation results in increased mitochondrial damage in brown adipose tissue (30). Less is known about the regulation of Trx-2 in response to a physiological challenge; however, Zhang et al. (31) have shown that 8 h of ambient hyperthermia (34°C) in broilers decreases hepatic Trx-2 gene expression.

Finally, Ref-1 is a redox-sensitive protein with several functions in the cell including DNA base excision repair, inhibition of hydrogen peroxide production, and stimulation of the AP-1 transcription factor (32–34). In cell culture systems, Ref-1 translocates to the mitochondria upon stimulation with hydrogen peroxide and phorbol 12 myristate 13-acetate (PMA), conferring protection to these stressors (35–37). Previously, we have shown that Ref-1 protein in the liver is elevated by aging, but that heat stress-induced expression of Ref-1 is impaired in old animals (38). It is unknown whether Ref-1 translocates to the mitochondria in our model of aging and hyperthermia.

Given the importance of these redox-sensitive proteins, it was of interest to investigate their regulation in this model of aging and heat stress. We hypothesized that mitochondria would demonstrate greater abundance of Sirt-3, MnSOD, Trx-2, and Ref-1 after heat stress, but that this response would be attenuated by aging. In this short report, we demonstrate evidence for mitochondrial translocation of these proteins and differential regulation by aging.

MATERIALS AND METHODS

All animal protocols were approved by the University of Iowa Institutional Animal Care and Use Committee, under Protocol No. 0606117. Male Fischer 344 rats (young: 6 mo, old: 24 mo) were obtained from the National Institute on Aging and housed in individual cages with access to food and water ad libitum. Animals were randomly divided into six groups: young, non-heated controls (YC); young, 2 h post-heat stress (Y2); young, 24 h post-heat stress (Y24); old non-heated controls (OC); old, 2 h post-heat stress (O2); and old 24 h post-heat stress (O24). Sample sizes are 5–6 animals per group except for the O24, in which n = 3. Animals exposed to heat stress underwent a two-heat stress protocol, mimicking what humans would experience during a multi-day heat wave, which are associated with morbidity and mortality in aged individuals (4–6). To familiarize the animals with the heating protocol, on each of the two preceding days before the first trial, animals were fitted with plastic temperature probes (Yellow Springs Instruments) placed 6–7 cm into the distal colon and were left in their plastic cage for ∼30 min. During each heat stress trial, rats were again fitted with the temperature probe that was connected to a digital temperature monitor. Animals were placed into a plastic cage, conscious and unrestrained. Using a heat lamp, colonic temperature (T_co) was elevated to 41°C over a period of 60 min and was then maintained at 41°C for an additional 30 min (Fig. 1). This protocol is similar to passive heat stress studies in humans that brought participants to the limit of thermotolerance in 1 h (39, 40). T_co was monitored continuously and recorded every 5 min. After each heat stress trial, animals were allowed to cool passively. Animals often try to remove the probe during heating; therefore, to reduce additional stress, only three time points were recorded after heating. A second heating trial was conducted 22.5 h after the first, and animals were euthanized at 2 and 24 h after the second heating trial. These times were chosen to coincide with a previous investigation demonstrating that oxidative stress is maximized at 2 h and subsides at 24 h (10). Control, nonheated animals were exposed to a sham heating protocol and were euthanized at similar times as the heated animals. No animals died before the conclusion of the study. Investigations from our laboratory have demonstrated that this repeated challenge elicits organ damage that is exaggerated in old, compared with young animals; hence, it is a physiologically relevant challenge that models what humans would experience during a heat wave.

Figure 1.

Colonic temperatures during the first (A) and second (B) heating trials. Temperatures were monitored continuously and recorded every 5 min. Each heating trial ended at 90 min, and temperatures were recorded 15 min into recovery. Means (+SD) temperatures are from young (circles) and old (squares) animals.

Animals were euthanized with an overdose of pentobarbital sodium (80 mg/kg) given intraperitoneally; their livers were excised and quickly rinsed in 0.9% saline, and mitochondrial fractions were obtained from freshly isolated livers, as previously described (11). Briefly, livers were minced with clean razor blades in isolation medium (110 mM mannitol, 20 mM sucrose, 2 mM Tris, 1 mM EDTA, 1% BSA, pH 7.4), and then homogenized with a Sonics Vibracell homogenizer in 10 volumes of isolation medium. The homogenate was centrifuged at 4°C for 10 min at 500 g and the supernatant was then centrifuged at 8,000 g for 10 min. Pellets were washed twice in storage medium (110 mM mannitol, 20 mM sucrose, 2 mM Tris, 1 mM EGTA), and then frozen for subsequent analysis. Frozen mitochondrial pellets were thawed, and then homogenized in lysis buffer (50 mM Tris, 150 mM NaCl, 0.25% sodium deoxycholate, 1% triton-X, 1 mM EDTA) with HALT protease and phosphatase inhibitor cocktail (1:100; Thermo Fisher #1861281) and the protein content of the lysates was analyzed via the Bradford protein assay. This mitochondrial isolation protocol resulted in a very pure fraction of mitochondria as demonstrated by the presence of mitochondrially localized proteins (MnSOD and Sirt-3), and the absence of cytosolic proteins (CuZnSOD and GAPDH; Supplemental Fig. S1A). Whole tissue liver lysates from a previous study (38) were utilized in confirmation studies on Sirt-3 and Ref-1 (as described in the results).

Equal amounts of mitochondrial protein (30 μg) were immunoblotted as previously described (38) using primary antibodies for MnSOD (Assay Designs # SOD-110; 1:5000), Cu/ZnSOD (Assay Designs # SOD-100;1:5000), Sirt-3 (Santa Cruz Biotechnologies # 99143; 1:500); acetylated lysines (Ac-K; Cell Signaling Technology 9441, 1:500); Ref-1 (Santa Cruz Biotechnologies, # 17774; 1:500), and Trx-2 (Cell Signaling Technology # 13322; 1:500). Following an overnight primary antibody incubation at 4°C in tris-buffered saline with tween (TBST) with 5% nonfat dry milk, membranes were washed with TBST, and then treated with either goat anti-rabbit (Santa Cruz Biotechnologies; sc-2030) or sheep anti-mouse (Amersham; NA931V) at dilutions of 1:4,000 in 5% milk TBST for 45 min at room temperature. Membranes were washed in TBST, and signals were developed using chemiluminescent substrate (Thermo Scientific Super Signal West Pico #34577). Images were taken with the Chemi-Doc XRS+ imaging system and Image Lab program (BioRad). Immunoblots displayed prominent bands at the appropriate molecular weights for each protein (Figs. 2, 4, and Supplemental Fig. S1). Where slightly different molecular weights are indicated on images, this is due to the use of different molecular weight standards. The brightness of each band of interest was quantified with the Image laboratory program. After probing for each protein, membranes were stained with Ponceau-S staining solution, which develops a red color dependent on the abundance of protein on the membrane. For each sample, the protein of interest was normalized to its own Ponceau-stained lane; this technique has been validated by multiple laboratories as a reliable and quantitative method (41, 42).

Figure 2.

Heat stress stimulates mitochondrial Sirt-3 accumulation and hyperacetylation of proteins. Representative blots (A) for Sirt-3, acetylated lysines (Ac-K), and Ponceau (Pon) in young and old control animals and at 2 and 24 h after heat stress. Quantitation of Sirt-3 (B), and Ac-K (C), normalized to the Ponceau stain. Results are expressed as means plus standard deviation with individual values shown. *Significant (P < 0.05) effect of age within the control or heated group. †Significant effect of heat stress within an age group. Significant main effects of age (P = 0.019) and heat stress (P = 0.002) were determined for Sirt-3. Significant main effects of age and heat stress were observed for Ac-K (P < 0.001).

Figure 4.

Evidence for heat-induced mitochondrial translocation of Ref-1. A: representative immunoblots for Ref-1 and Ponceau (Pon) with the quantitation beneath. Results are expressed as means plus standard deviation with individual values shown. †Significant effect of heat stress within an age group (main effect of heat stress, P < 0.001). In B, the molecular weights of Ref-1 from mitochondria and whole tissue are displayed. Lanes are shown in their entirety, with the Ponceau-stained membrane underneath. We loaded greater amounts of protein in the tissue lysates (60 µg) to match the signal intensity in the mitochondrial samples (30 µg). Mitochondrial (m) Ref-1 is from 2 h heated young or old animals. Whole tissue (wt) Ref-1 is from whole liver lysates from nonheated young and old animals. Mitochondrial (m) Ref-1 was consistently ∼10 kDa heavier than whole tissue (wt) Ref-1. Results are representative of five samples in each of the four groups. Note the greater abundance of Ref-1 in whole tissue samples from aged rats, which we reported previously (38).

The normalized densities of proteins for each group were further normalized to the young, nonheated control group, which was given a value of 1. Each comparison (i.e., young vs. old within each heating condition, and control vs. heated in each age group) was performed on the same gel with three to six samples in each group. Experiments were conducted until all comparisons within each heating and age group were made and immunoblots were run in triplicate.

Statistics

The normality of each data set (YC, Y2, Y24, OC, O2, O24, for each of the five protein targets) was determined with a Shapiro-Wilk test. The only set that was not normally distributed was the Ref-1, Y24 group. Thus, a nonparametric test (Mann–Whitney) was used when comparing YC versus Y24 and Y24 versus O24 within this data set. All other data were analyzed with a two-factor ANOVA (factors: age and heat stress) with IBM SPSS software, version 29. When significance was initially indicated by the ANOVA (P value less than 0.05), post hoc analyses were conducted using the Bonferroni post hoc test. Results were considered significant at P values less than 0.05. Results are presented in figures as means + standard deviation, with individual values indicated as circles (young animals) or squares (old animals).

RESULTS

Figure 1 shows T_co during the first (A) and second (B) heating trials. Heating rates (∼0.06°C/min, ±0.005 SD) were identical during trials, and the temperatures of nonheated animals fluctuated between 37.0°C and 38.5°C [not shown and (43)]. The greater initial colonic temperatures in young animals were likely due to more ambulatory activity in this age group, which was qualitatively observed. After the initial 15 min, young and old animals experienced identical heating rates. Temperatures from three time points (95, 100, and 105 min) into recovery were collected, which demonstrated an immediate decrease in T_co after heating at a rate of ∼0.09°C/min (±0.014 SD).

Sirt-3 and Acetylated Lysine Residues

We observed significant main effects of age (P = 0.019) and heat stress (P = 0.002) on mitochondrial Sirt-3 protein (Fig. 2, A and B). In the control condition, Sirt-3 protein was 35% greater in old, compared with young animals (P = 0.025). This age-related difference was confirmed in whole-tissue liver lysates from a separate set of animals (n = 7 per group), in which Sirt-3 was 31% greater in aged livers (Supplemental Fig. S1B). In mitochondrial fractions, young rats demonstrated greater Sirt-3 protein at 2 and 24 h after heat stress, compared with nonheated values (P = 0.010 and P = 0.002 respectively). In old animals, heat stress resulted in a significant increase in Sirt-3 only at the 2 h timepoint (P = 0.05).

Acetylated lysine residues (Ac-K) were detected as two prominent bands which occurred as a doublet just below 40 kDa (Fig. 2A). Given their proximity, we quantified these bands together. Fainter bands were observed at other molecular weights, but they were less consistent and therefore not quantified. The molecular weights of these acetylated proteins are consistent with a previous investigation in hepatic mitochondria (21). Significant main effects of age (P < 0.001) and heat (P < 0.001) were determined. Acetylated lysines in young animals were significantly greater than in old animals in the control (P = 0.044), 2 h (P = 0.020), and 24 h groups (P = 0.012). The observation that old animals had ∼30% less Ac-K than their younger counterparts matches results from a previous investigation from another laboratory (44). Heat stress was associated with a significant increase in Ac-K in the young 24 h group only (P = 0.002; Fig. 2C).

Mitochondrial Antioxidant Protein Abundance

Although MnSOD protein amounts were similar between young and old animals at all time points, we observed a significant main effect of heat stress on MnSOD (P = 0.003; Fig. 3, A and B). In both age groups, we observed a subtle, but significant increase in MnSOD protein at 24 h after heat stress compared with controls (young, P = 0.010; old, P = 0.009; Fig. 3, A and B). We observed a significant age-by-heat interaction effect in Trx-2 (P = 0.003) with old values declining significantly at 2 h after heat stress (P = 0.006). At the 2 h time point, Trx-2 protein abundance in the young animals was significantly greater than in the old animals (P = 0.001; Fig. 3, A and C).

Figure 3.

Mitochondrial abundance of manganese superoxide dismutase (MnSOD) and thioredoxin-2 (Trx-2) in young and old rats after heat stress. Representative blots for MnSOD and Trx-2 (A), with the Ponceau-stained membrane underneath (Pon). Quantitation of MnSOD (B) and Trx-2 (C), normalized to the Ponceau stain. Results are expressed as means plus standard deviation with individual values shown. *Significant (P < 0.05) effect of age within the control or heated group. †Significant effect of heat stress within an age group (MnSOD, main effect of heat stress, P = 0.003; Trx-2, significant age-heat stress interaction, P = 0.003).

Redox Factor-1

Mitochondrial Ref-1 protein was barely detectable under control conditions in both young and old animals (Fig. 4A), and no significant main effect of aging was observed. However, a significant main effect of heat stress was observed (P < 0.001). In young animals at 2 h after heat stress, we observed a sixfold elevation in mitochondrial Ref-1 protein expression compared with control values (P = 0.008). In old rats, Ref-1 protein abundance at 2 h was ninefold higher than the control condition (P = 0.031). Ref-1 protein abundance at 24 h was not different than control values in either age group (Fig. 4A). The mitochondrially localized Ref-1 (from the 2 h samples) consistently migrated at a higher molecular weight (∼10 kDa greater) than nonheated control samples from whole tissue (Fig. 4B), in which Ref-1 is predominantly nuclear (38). We did not observe such differences in molecular weight in the other proteins evaluated.

DISCUSSION

Since stress resistance declines with aging, it is important to identify specific cellular pathways that might be targeted to lessen stress-induced morbidity and mortality, especially given the current increasing trend of global temperatures. Interestingly, our results showing similar mitochondrial accumulation of Sirt-3, MnSOD, and Ref-1 between young and old animals after heat stress demonstrate that some aspects of the cellular stress response remain intact with aging. Furthermore, because the mitochondrial genome codes exclusively for proteins of the electron transport chain, our results suggest that Sirt-3, MnSOD, and Ref-1 translocate to the mitochondria after hyperthermia in vivo. While we did not observe significant heat-induced accumulation of Trx-2 in young rats, it declined in the old rats at 2 h after heat stress, revealing another potential mechanism for age-related mitochondrial dysfunction.

Previous investigations on Sirt-3 expression in the aged liver have yielded differing results. For example, Kwon and associates demonstrated less Sirt-3 protein in male 20-mo-old, compared with 6-mo-old C57BL/6J mice (45). These mice develop inflammation more readily compared with the Fischer 344 rat (46), and inflammation decreases Sirt-3 protein abundance (47, 48), which could explain the differences between our observations. In C57BL/6 mice, there were similar levels of Sirt-3 protein between 2- and 22-mo-old animals, but the authors did not report the sexes utilized (49). Skeletal muscle in women has greater Sirt-3 expression than men (50), and hepatic Sirt-3 in male 129/SV mice declines after a chronic high-fat diet while Sirt-3 in female mice is maintained (51). Therefore, if the previous study (49) investigated female animals, it is possible that sex differences contributed to similar Sirt-3 protein with age. Finally, a study in male Fischer 344 rats observed no differences in hepatic Sirt-3 protein abundance between young and old animals (44). Thus, it was surprising to observe that Sirt-3 was elevated in liver mitochondria from aged animals in our study. To confirm this observation, we measured Sirt-3 in whole-tissue lysates from a completely different sample set of young and old Fischer 344 rats and again observed that aging significantly elevated hepatic expression of Sirt-3. This is consistent with the observations in the rat hippocampus (52) as well as the heart and skeletal muscle of mice (53). Given that Nakamura et al. (44) investigated Sirt-3 expression in 9- and 27-mo-old Fischer 344 rats (we evaluated 6- and 24-mo-old animals), it is possible that Sirt-3 either increased between 6 and 9 mo, or decreased between 24 and 27 mo. Thus, a more thorough time course investigation on male and female animals is needed to reconcile these results.

Stressors such as endoplasmic reticulum stress and bile duct ligation increase Sirt-3 protein expression in the liver (15). To our knowledge, this is the first demonstration of heat stress-induced Sirt-3 accumulation in mitochondria in vivo. Our results demonstrating hyperacetylation of mitochondrial proteins after heat stress could suggest that Sirt-3 activity does not increase enough to prevent hyperacetylation of mitochondrial proteins, or that nonenzymatic acetylation is enhanced. While hepatic protein hyperacetylation has been observed in response to a high-fat diet (20, 54), ethanol intoxication (21), and treatment with ferric nitrilotriacetate (55), to our knowledge, this is the first demonstration of elevated protein acetylation after hyperthermia. One commonality among these stressors is that they promote a pro-oxidative cellular and mitochondrial milieu. Importantly, our heat stress model is associated with an elevation in hydrogen peroxide (10), which directly modifies lysine residues and results in the production of acetylated proteins in isolated chemical reactions (55). Thus, enhanced production of hydrogen peroxide after heat stress could mediate the formation of acetylated proteins within the mitochondria. We note that aged animals had greater expression of Sirt-3, along with lower Ac-K in the control condition, and a nonsignificant increase in Ac-K after heat stress. While this is the expected relation between Sirt-3 and Ac-K, we cannot conclude causation with the experiments performed here. Changes in acetylation in the absence of changes in hepatic Sirt-3 protein can occur (20, 21, 44), along with nonenzymatic alterations in Ac-K concentrations due to changes in liver acetyl-CoA concentrations (56). Thus, future studies that directly modulate Sirt-3 protein and/or activity will be needed to establish its effect on Ac-K in this model.

The heat-induced decrease in Trx-2 protein in old animals is noteworthy because it has implications for mitochondrial permeability. For example, overexpression of Trx-2 in mice prevents hepatic mitochondrial permeability due to tert-butylhydroperoxide [an oxidant; (57)]. In addition, Trx-2 protects against hepatotoxicity, including mitochondrial cytochrome C release after diquat treatment (29). Thus, the age-related enhancement of mitochondrial cytochrome C release after heat stress observed previously in this model (11) could be due to the decline in Trx-2 abundance. Other studies from our laboratory utilizing the same experimental heat stress model have demonstrated excessive ultrastructural hepatic mitochondrial damage and less ATP in old, compared with young rats at similar timepoints after heat stress (12). Taken together, our current results suggest that the decline in Trx-2 after heat stress leads to compromised mitochondrial function in old animals. The loss of Trx-2 versus the gain in other proteins is an interesting phenomenon worthy of future investigation. Compared with the other proteins investigated here, Trx-2 is a relatively small molecular weight protein (∼12 kDa); thus, given the greater degree of structural damage in aged mitochondria (11, 12), Trx-2 would be more likely to escape the mitochondria, compared with higher molecular weight proteins. This is consistent with the loss of cytochrome C in this model (11), which also weighs 12 kDa. Evidence exists for differential entry and exit of hepatic mitochondrial proteins after heat stress (11, 58) although the specific mechanisms remain unclear.

We and others have shown that Ref-1 is induced in vivo by stressors such as hyperthermia in the liver (38) and acute exercise in skeletal muscle (59), both of which elevate hydrogen peroxide production (10, 59). Mitochondrial translocation of Ref-1 has been observed in response to hydrogen peroxide in cell culture models (35–37), yet to our knowledge, this is the first demonstration of stress-induced Ref-1 mitochondrial translocation in vivo. Interestingly, this response was not impaired with aging, and the robust accumulation of Ref-1 in young and old animals is noteworthy and novel. Since old animals exhibit greater oxidative stress and injury after heat stress (10), the functional implications of similar Ref-1 translocation will be important to delineate in future studies. Given its function as a DNA repair enzyme, it is likely that Ref-1 translocation is a response to damaged or oxidized mitochondrial DNA. Finally, we observed a difference in the molecular weights between mitochondrial Ref-1 (∼46 kDa) and Ref-1 from whole tissue [∼36 kDa, which is predominately nuclear in liver; (38)], which is a unique finding. This greater molecular weight of mitochondrial Ref-1 likely suggests a post-translational covalent modification, which will be important to identify in future investigations.

Despite showing several novel results, our study has some limitations. First, we evaluated two timepoints after the second heat stress. These timepoints were chosen to match a previous investigation from our laboratory showing that oxidative stress is maximized at 2 h and subsides at 24 h after the second heating (10). A more thorough time course, including timepoints after 24 h, would be useful to perform in a more comprehensive investigation. Second, we note the sample size of 3 in the aged 24-h group. Due to constraints in our budget and other resources, we were unable to evaluate additional animals in this group. Finally, we evaluated a select number of protein targets in the mitochondria, and we acknowledge that this is not a comprehensive evaluation of mitochondrial antioxidant proteins. In particular, Trx-2 and peroxiredoxins 3 and 5 interact to prevent oxidative stress in the mitochondria; therefore, it will be important to investigate this class of proteins going forward.

Overall, we have demonstrated that old rats manifest a response to physiological stress that, in some respects, is similar to young animals. Given the decline in Trx-2 protein, and that there is an overall deleterious response to heat stress in the old animals, future studies should focus on increasing Trx-2 protein or activity to improve outcomes in this age group. Further unraveling the cellular responses to hyperthermia will be important as our planet continues to warm.

DATA AVAILABILITY

Data are available from the corresponding author upon reasonable request.

SUPPLEMENTAL DATA

Supplemental Fig. S1: https://doi.org/10.6084/m9.figshare.24256645.

GRANTS

K.C.K. was supported by National Institutes of Health Grant AG-12350. S.A.B. was supported by a Faculty Development Grant from Penn State Abington.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.L.H., K.C.K., and S.A.B. conceived and designed research; J.L.H. and S.A.B. performed experiments; J.L.H., K.C.K., and S.A.B. analyzed data; J.L.H., K.C.K., and S.A.B. interpreted results of experiments; S.A.B. prepared figures; S.A.B. drafted manuscript; J.L.H., K.C.K., and S.A.B. edited and revised manuscript; J.L.H., K.C.K., and S.A.B. approved final version of manuscript.