Muscular apoptosis but not oxidative stress increases with old age in a long-lived diver, the Weddell seal

Diving vertebrates employ a physiological diving response, limiting oxygen consumption by peripheral tissues to conserve oxygen for the hypoxia-sensitive central nervous system during a dive. This response is characterized by apnea, bradycardia and peripheral vasoconstriction; the last of these is generally tissue specific and potentially profound (Kooyman and Campbell, 1972; Meir et al., 2008; Ponganis et al., 2008; Zapol et al., 1979). As an animal surfaces at the conclusion of a dive, peripheral tissues are reperfused with oxygenated blood (Elsner et al., 1998). In terrestrial mammals, cycles of ischemia–reperfusion (IR) and the associated hypoxia–reoxygenation result in cellular, tissue and organ damage (McCord, 1985; Salvadori et al., 2015). These insults are typically attributed to the increased generation of reactive oxygen species (ROS) by xanthine oxidase, NADPH oxidases and mitochondria (Granger and Kvietys, 2015). In contrast, the catastrophic IR injuries observed in terrestrial species do not occur in naturally hypoxia-tolerant vertebrates such as seals, penguins and turtles (Fago and Jensen, 2015; Kooyman and Ponganis, 1998; Vázquez-Medina et al., 2012; Zenteno-Savín et al., 2002). The degree to which damage may manifest and accumulate in these animals as a function of age progression, however, remains unknown.

In addition to IR injury, ROS are implicated in the aging process. First described by Harman in 1956, the free radical theory of aging purports that increased ROS generation – and the resulting accumulation of cellular oxidative damage – contributes to several pathologies and a decline in physiological function associated with advancing age (Harman, 1956). While the specific mechanisms leading to oxidative damage in aging tissues are multiple, an overall imbalance in redox metabolism drives structural and functional changes in aging organisms (Sohal and Orr, 2012). Both oxidative stress and apoptosis are associated with loss of muscle mass (sarcopenia) in aging rodents and humans (Dirks and Leeuwenburgh, 2002; Kannan and Jain, 2000; Kujoth et al., 2005; Lee and Wei, 2007; Leeuwenburgh, 2003; Song et al., 2006). Wild vertebrates are considered to be largely exempt from these processes, as any negative impact on foraging or predator evasion would lead to mortality by extrinsic forces prior to the accumulation of lethal cellular damage in an individual (Kirkwood and Austad, 2000). However, cellular senescence and aging may manifest in long-lived mammals, including seals. Mixed evidence exists for aging in pinnipeds: reproductive senescence has been documented in subantarctic fur seals (Beauplet et al., 2006), but not southern elephant seals (Pistorius and Bester, 2002). Based on our prior observations of skeletal muscle remodeling in old Weddell seals (Hindle et al., 2009), we assessed whether apoptosis is associated with aging and changes in redox metabolism in these same animals.

The cellular processes underlying morphological changes in the skeletal muscle of long-lived diving vertebrates have not been investigated. Therefore, we tested the hypothesis that increased oxidative stress is related to apoptotic cell death and the previously observed morphological changes in aging Weddell seal skeletal muscles. We evaluated antioxidant enzyme activities and expression as well as markers of lipid peroxidation and apoptosis in Weddell seal locomotor (longissimus dorsi) and maneuvering (pectoralis) muscles, which are anticipated to experience differential ROS generation during diving as a result of muscle-specific activity patterns (Kanatous et al., 1999; Powers et al., 2011). Seals were categorized as either peak reproductive (‘young’, aged 9–16 years) or post-peak reproductive (‘old’, aged 17–26 years) animals to determine whether oxidative stress is associated with previously observed age-associated remodeling of contractile and connective tissue in skeletal muscle. We also hypothesized that known behavioral and activity differences between male and female Weddell seals (i.e. increased fighting in males) are associated with increased oxidative stress as has been observed in other phocids (Sharick et al., 2015).

Biopsies were collected from 49 free-ranging adult Weddell seals, Leptonychotes weddellii (Lesson 1826) (n=26 male, n=23 female; age 9–26 years) in Erebus Bay, McMurdo Sound, Antarctica (77°55′S, 166°40′E) during two consecutive breeding seasons (October–December 2006, 2007). Before capture, animal age (to the nearest year) and sex were visually determined by use of pre-existing flipper tags (Cameron and Siniff, 2004). Only seals appearing to be in good health were selected for sampling; a health and condition assessment for the study animals is reported elsewhere (Mellish et al., 2011). Pregnant or lactating females were excluded from sampling. Pregnancy status of females was assessed by transabdominal ultrasound (Sonosite 180 Vet; Bothwell, WA, USA). Lactation status was determined from an existing database of female seals documented as having produced a pup in that season. Seals were first restrained using a head-bag (Stirling, 1966), then chemically sedated via intramuscular injection of ketamine hydrochloride combined with either diazepam (2 mg kg⁻¹ ketamine:0.01 mg kg⁻¹ diazepam) or midazolam (2 mg kg⁻¹ ketamine:0.1 mg kg⁻¹ midazolam). We have previously found that this is a reliable and effective field cocktail to chemically immobilize Weddell seals for procedures involving tissue biopsy (Mellish et al., 2010). Animals were weighed using a suspended force transduction scale (San Diego Scale Company, San Diego, CA, USA; accurate within ±0.5 kg). Muscle biopsies were collected from longissimus dorsi and pectoralis muscles under anesthesia as previously described (Hindle et al., 2009), flash frozen in liquid nitrogen and stored at −80°C until analysis. All work was conducted under permits from the Marine Mammal Protection Act (#1034-1854) and Antarctic Conservation Act (#2007-007) and in accordance with institutional animal use protocols (Texas A&M University #2006-160; Alaska SeaLife Center #05-004; Oregon State University #3454).

Longissimus dorsi and pectoralis muscle samples were homogenized separately with a glass-on-glass tissue grinder in ice-cold lysis buffer [10 mmol l⁻¹ Hepes pH 7.5, 350 mmol l⁻¹ NaCl, 20% glycerol, 1% Igepal-CA630, 1 mmol l⁻¹ MgCl₂, 0.1 mmol l⁻¹ DTT, protease inhibitor tablet (11836170001, Roche Applied Science, Indianapolis, IN, USA)] at 1:20 w/v. Homogenates were centrifuged at 12,000 g to remove cellular debris and nuclear material. The supernatant was divided into aliquots and stored at −80°C until analyses, at which time the homogenate was further diluted in appropriate reaction buffers as detailed below. The protein content of muscle homogenates was measured using a modified Bradford assay (kit 23236, Pierce, Rockford, IL, USA). All enzyme activity assays were conducted at room temperature in triplicate, and were batch processed to limit between-day reagent variability.

Catalase (EC 1.11.1.6) activity of skeletal muscle homogenates was assayed following the methods of Aebi (1984) and Cohen et al. (1970). Homogenates (1:20 w/v) were first diluted 1:4.5 in ice-cold 100 mmol l⁻¹ KPO₄ without EDTA. Homogenates were then incubated (10:1 v/v) with ethanol for 30 min on ice to decompose existing inactive catalase–H₂O₂ complexes in the crude extract (Cohen et al., 1970). Next, homogenates were incubated with 1% Triton X-100 for 15 min at room temperature to improve enzyme accessibility for the assay, as catalase is typically localized in the peroxisomes (Cohen et al., 1970). The reaction was initiated by the addition of 10 mmol l⁻¹ hydrogen peroxide. Catalase activity was assayed by directly monitoring H₂O₂ extinction at 240 nm (Aebi, 1984). Blank runs were conducted in the absence of homogenate. One unit of activity was defined as the amount of enzyme required to reduce 1 µmol of H₂O₂ per minute. Glutathione reductase (GR, EC 1.6.4.2) catalyzes the NADPH-dependent reduction of glutathione (GSSG). GR activity was measured by tracking depletion of NADPH at 340 nm for 3 min in a solution containing skeletal muscle homogenate (1:700 to 1:11,000 w/v) and reaction buffer (0.15 mmol l⁻¹ NADPH, 1.4 mmol l⁻¹ GSSG, 10 mmol l⁻¹ EDTA in 100 mmol l⁻¹ potassium phosphate buffer, pH 7.4) according to Carlberg and Mannervik (1985). Blanks were run in the absence of homogenate. One unit of activity was defined as the amount of enzyme required to reduce of 1 µmol of NADPH per minute. Glutathione peroxidase (GPx, EC 1.8.1.7) activity was measured according to Flohé and Günzler (1984). In brief, reaction cocktail [0.3 U ml⁻¹ GR, 1.25 mmol l⁻¹ glutathione (reduced form, GSH), 0.2 mmol l⁻¹ NADPH, 10 mmol l⁻¹ EDTA in 100 mmol l⁻¹ potassium phosphate buffer, pH 7.4] and muscle homogenate (1:5000 w/v) were incubated for 3 min at room temperature; the reaction was initiated with 12 mmol l⁻¹ t-butyl hydroperoxide and monitored at 340 nm over 4 min. Blank runs were conducted in the absence of homogenate. One unit of activity was defined as the amount of enzyme required to oxidize 1 µmol of NADPH per minute. Glutathione S-transferase (GST, EC 2.5.1.18) activity was monitored at 340 nm to track the formation of thioether product resulting from the reaction between GSH and 1-chloro-2,4-dinitrobenzone (Habig and Jakoby, 1981). Homogenates (1:1350 w/v) were incubated for 4 min with reaction buffer (1 mmol l⁻¹ GSH, 60 mmol l⁻¹ EDTA in 100 mmol l⁻¹ potassium phosphate buffer, pH 6.5) and 1 mmol l⁻¹ CDNB. Blanks were run in the absence of homogenate. One unit of activity was defined as the amount of enzyme required to yield 1 µmol of reaction product per minute. All enzyme activities were normalized to total protein content. The activity of isocitrate dehydrogenase (ICDH, EC 1.1.1.42) was assayed following the method of Cleland et al. (1969). This assay does not distinguish ICDH activity arising from cytosolic and mitochondrial isoforms; however, in skeletal muscle, ICDH is expected to be predominantly localized in the mitochondria (Plaut et al., 1983). Muscle homogenate (1:10,000 w/v) was combined with reaction cocktail (5.6 mmol l⁻¹ isocitrate, 11 mmol l⁻¹ MnCl₂ in 100 mmol l⁻¹ Tris buffer, pH 7.4). The reaction was initiated with NADP (0.6 mmol l⁻¹) and the rate of NADPH production was followed for 0–3 min at 340 nm. Blank runs were conducted without isocitrate. One unit of activity was defined as the amount of enzyme required to form 1 µmol of NADPH per minute.

The relative abundance of copper zinc superoxide dismutase (CuZnSOD, SOD-1, cytosolic, EC 1.15.1.1), manganese superoxide dismutase (MnSOD, SOD-2, mitochondrial, EC 1.15.1.1) and HSP70 was measured in tissue homogenates by western immunoblotting. Skeletal muscle homogenate (20 µg protein) was separated by SDS-PAGE and transferred to nitrocellulose membrane (Bio-Rad, Hercules, CA, USA). Consistent loading and transfer were verified by Ponceau S staining. Membranes were rinsed with PBS and blocked at room temperature for 4 h in 5% non-fat milk (in PBS with 0.1% Tween-20). After blocking, membranes were washed (3×5 min) with PBS and 0.4% Tween-20, and incubated overnight at room temperature with primary antibodies in PBS as follows: CuZnSOD, 1:500 (rabbit polyclonal, Santa Cruz Biotechnology FL-154, Santa Cruz, CA, USA); MnSOD, 1:5000 (rabbit polyclonal, Stressgen SOD-111, Ann Arbor, MI, USA); HSP70, 1:10,000 (rabbit polyclonal, Stressgen SPA-812). After incubation, membranes were washed (3×5 min) with PBS and 0.4% Tween-20 and incubated with HRP-conjugated anti-rabbit secondary antibody (Rockland Immunochemical, Gilbertsville, PA, USA) for 90 min at room temperature. Blots were visualized with enhanced chemiluminescence detection (Supersignal West Pico Chemiluminescent Substrate, Pierce, Rockford, IL, USA) on a CCD digital imager. Relative density [area×(density−background for each lane)] of grayscale images was calculated with ImageJ software (v1.37s, National Institutes of Health, USA). Densities were normalized to total protein.

Lipid hydroperoxide levels were quantified using the Xylenol Orange oxidation assay (Hermes-Lima et al., 1995). Reaction constituents (225 µl 1 mmol l⁻¹ FeSO₄, 90 µl 0.25 mol l⁻¹ H₂SO₄, 90 µl 1 mmol l⁻¹ Xylenol Orange, 468 µl 0.055 mol l⁻¹ H₂SO₄, 180 µl homogenate diluted 1:50 w/v) were prepared in 100 mmol l⁻¹ potassium phosphate buffer at pH 7.4 and incubated in the dark for 12 h. Total hydroperoxides were calculated based on a concurrently run t-butyl hydroperoxide standard curve at 580 nm. Lipid peroxidation levels were also quantified by 4-hydroxynonenal-Lys (4-HNE-Lys) detection using western immunoblotting. Homogenate (30 µg protein) was separated by SDS-PAGE, then transferred onto nitrocellulose membrane (Bio-Rad). Consistent loading and transfer were verified by Ponceau S staining, then blots were rinsed in PBS and blocked with 5% non-fat milk (in PBS with 0.1% Tween-20) at room temperature for 4 h. Blocked membranes were washed (3×5 min) with PBS and incubated overnight at room temperature with primary antibody (rabbit polyclonal, 1:500, Calbiochem 393206, San Diego, CA, USA). Membranes were washed (3×5 min with PBS) and incubated with HRP-conjugated anti-rabbit secondary antibody (Rockland Immunochemical, Gilbertsville, PA, USA) for 90 min at room temperature. Blots were visualized and analyzed as above.

Apoptotic cell death was measured by an ELISA assay based on the detection of the cytosolic presence of mononucleosomes and oligonucleosomes characteristic of nuclear breakdown (Cohen, 1999; Duvall and Wyllie, 1986). The assay was performed in accordance with the manufacturer's instructions (Roche 11544675001). Data were normalized to the background absorbance for each plate individually, and results are reported as arbitrary absorbance units (AU) per μg total protein.

Samples for all cohorts (young, old, male, female) were collected during both sampling years. Sample numbers were determined by federal permitting constraints on sampling a protected species. All assays were run in duplicate at the Albert P. Crary Science and Engineering Center at McMurdo Station, Antarctica. Each muscle biopsy site was considered separately for analysis. Batch differences between sampling years were corrected by scaling data from the first year by the ratio of the year means. Within muscle site-specific datasets, outliers for each assay were identified and excluded as points lying beyond 1.5× interquartile range (whiskers) for each cohort and sex. Data were tested using the Shapiro–Wilk test to determine whether they were normally distributed. For normally distributed data (Shapiro–Wilk P≥0.05), variables were compared with an ANOVA model that included cohort, sex and cohort×sex as cofactors. For non-normally distributed data (Shapiro–Wilk P<0.05), variables were compared using the Scheirer–Ray–Hare test with cohort, sex and cohort×sex as cofactors. In only one case were cohort×sex interactions significant. Here, a post hoc Tukey HSD test was performed. For all tests, significance was set at α=0.05. Statistical analyses were performed using RStudio (version 1.1.423, Boston, MA, USA). All summary values are reported as means±s.e.m. Fold-change is reported as the ratio of compared means.

Biopsies were collected from 49 adult Weddell seals. Sample sizes for some assays were limited by the quantity of tissue collected, as indicated by distinct n values for each analysis. Age cohorts were delimited as ‘young’ (age 9–16 years, n=26 animals) and ‘old’ (age 17–26 years, n=23 animals) to correspond with our prior studies (Hindle et al., 2009), and were based on age differences in reproductive rates for this population of Weddell seals (Proffitt et al., 2007). Mass was 397.6±10.0 kg (range: 317.3–553.2 kg) for the young cohort and 423.5±11.4 kg (range: 292.3–508.2 kg) for the old cohort. Mass was 396.3±10.4 kg (range: 292.3–553.2 kg, n=26) for all males and 423.8±10.8 kg (range: 317.3–512.7 kg, n=23) for all females.

In old seals, the apoptotic index of longissimus dorsi (swimming muscle) was nearly double the level detected at the same site in young animals; this difference was significant (Fig. 1A). However, no significant differences between age cohorts were detected in either 4-HNE-Lys or lipid hydroperoxide levels (Fig. 1B,C). Notably, there were no age-associated differences in antioxidant profiles (Fig. 2, Table 1). Taken together, these data suggest that the apoptosis observed in the longissimus dorsi of old Weddell seals is not driven by increased oxidative stress.

In contrast to the longissimus dorsi, the non-swimming pectoralis muscle apoptotic index did not differ between age groups (Fig. 1A). However, there were several differences in markers of lipid peroxidation and antioxidant profiles between pectoralis from old and young adult Weddell seals. 4-HNE-Lys levels were 1.5 times more abundant in the old versus the young cohort (Fig. 1B). No differences were observed for lipid hydroperoxide levels (Fig. 1C). GPx activity was 1.5-fold higher in old versus young seals (Fig. 2). No age-associated changes were detected in pectoralis for other antioxidants assayed (Table 1). The age-specific lipid peroxidation and antioxidant profiles observed in pectoralis suggest that this tissue may incur limited oxidative damage as seals age, although this damage does not progress to apoptotic cell death. Here, the patterns of apoptosis, lipid peroxidation and antioxidant expression associated with advancing age differed between longissimus dorsi and pectoralis.

Apoptotic index, 4-HNE-Lys levels and lipid hydroperoxide levels did not vary with sex in longissimus dorsi (Fig. 3). GR activity was 1.4-fold higher in females than in males (Fig. 4A), although this difference was driven by decreased GR activity in old males (cohort×sex P=0.0498). Males displayed increased GPx activity (1.3-fold; Fig. 4B) and HSP70 expression (1.4-fold; Fig. 5) compared with females. GST activity, however, did not vary with sex (Fig. 4C), nor did MnSOD and CuZnSOD expression or the activities of catalase and ICDH (Table 2). These data suggest that male and female Weddell seals utilize sex-specific patterns of glutathione-related antioxidant enzymes to cope with oxidative stress in the primary locomotor muscle, the longissimus dorsi.

Apoptotic index, 4-HNE-Lys levels and lipid hydroperoxide levels did not vary with sex in pectoralis (Fig. 3). Activities of glutathione-related enzymes did not change with sex (Fig. 4), nor were there sex differences in MnSOD or CuZnSOD expression or the activities of catalase and ICDH (Table 2). The consistency in antioxidant profiles and lack of difference in oxidative damage across sexes for the pectoralis suggest that male and female seals do not experience differential oxidative stress in this tissue.

Seals are exposed to repeated bouts of IR associated with diving (Elsner et al., 1998). In terrestrial vertebrates, IR events increase ROS generation, oxidative stress, cell death and tissue damage (Kalogeris et al., 2012; McCord, 1985; Salvadori et al., 2015). Dysregulated ROS generation and accumulation of oxidative damage are also associated with cellular senescence and aging (Colavitti and Finkel, 2005). Therefore, long-lived diving vertebrates such as seals may potentially be exposed to oxidative damage accumulation that could impair cellular function. Phocid seal muscles have the capacity to generate ROS (Zenteno-Savín et al., 2002), and this capacity increases with post-natal maturation (Vázquez-Medina et al., 2011a), but it is not known whether this contributes to aging in these animals. Our previous work documented morphological muscular senescence in old Weddell seals without functional changes in the animals' diving ability (Hindle et al., 2009, 2011). Here, we show that senescence and apoptosis are not associated with increased oxidative stress in Weddell seal skeletal muscles. Furthermore, we found that males and females display differential antioxidant strategies and that redox metabolism differs between muscle types.

Oxidative stress contributes to senescence, apoptosis and muscular atrophy in rodents and humans (Ábrigo et al., 2018; Kujoth et al., 2005; Lee and Wei, 2007). Atrophy and functional impairment in a wild animal can negatively impact fitness, inducing mortality through indirect means. There was no association between markers of oxidative stress and apoptosis in skeletal muscles of Weddell seals. In the longissimus dorsi, apoptosis but not lipid peroxidation increased with advancing age, and there were no differences in lipid peroxidation or apoptosis between sexes. In the pectoralis, elevated 4-HNE-Lys levels in old seals were not associated with increased apoptosis, and there were no sex differences in either apoptosis or lipid peroxidation. In rats, aging-related apoptotic signaling in skeletal muscle decreases with physical activity (Song et al., 2006). The physical activity which defines the Weddell seal's life history may constrain increases in apoptosis in skeletal muscle, limiting functional changes.

Diving vertebrates restrict peripheral perfusion during dives, including to skeletal muscle (Guyton et al., 1995; Williams et al., 2011; Zapol et al., 1979). During such vasoconstriction, IR and hypoxia–reoxygenation cycles potentially expose tissues to oxidative injury. High concentrations of myoglobin in pinniped skeletal muscle support substantial oxygen stores in this tissue (Castellini and Somero, 1981; Kanatous et al., 1999, 2002), potentially buffering the dramatic changes in tissue oxygen concentrations that typically drive oxidant generation and reperfusion injury in non-diving mammals. Our observation that age-associated oxidative damage occurs in pectoralis (maneuvering muscle) but not in longissimus dorsi (swimming muscle) is consistent with higher concentrations of myoglobin in Weddell seal longissimus dorsi compared with pectoralis (Hindle et al., 2011; Kanatous et al., 2002).

Skeletal muscle fiber type composition may also help limit oxidative damage in seals. Muscles of deep-diving seals lack fast glycolytic (type IIB) fibers; oxidative fibers (type I and type IIA) dominate skeletal muscle composition (Kanatous et al., 2002, 2008; Moore et al., 2014). In Weddell seals, slow oxidative (type I) fibers comprise the majority of the swimming muscle (longissimus dorsi), while the non-swimming muscle (pectoralis) consists mainly of fast oxidative (type IIA) fibers (Kanatous et al., 2002). In aging – and often sedentary – humans and rodents, type II muscle fibers are more susceptible to sarcopenia (Holloszy et al., 1991; Nilwik et al., 2013), though this is apparently not reflected in physically active wild populations (Hindle et al., 2009, 2010). Studies in rodents show that type II fibers are also particularly susceptible to reperfusion injury (Chan et al., 2004; Walters et al., 2008), consistent with our observation that pectoralis but not longissimus dorsi accumulates oxidative damage in old animals. In Weddell seals, 4-HNE-Lys levels increased with age in pectoralis, which is composed primarily of type IIA fibers. This increase was accompanied by an increase in GPx activity, which may represent a response that helps protect against apoptosis in pectoralis. Life-long physical activity in rodents limits sarcopenia, oxidative stress and apoptosis while preserving expression of heat shock proteins and superoxide dismutase (Kim et al., 2008, 2015). Physically active free-ranging wild populations such as Weddell seals may benefit similarly.

Endogenous antioxidants are key to managing oxidative stress in mammals. Various marine mammal species exhibit high antioxidant levels both in circulation and within tissues, including skeletal muscle, compared with non-diving mammals (García-Castañeda et al., 2017; Vázquez-Medina et al., 2006; Vázquez-Medina et al., 2007; Wilhelm Filho et al., 2002). Moreover, recent evidence shows positive selection of antioxidant genes in diving mammals (Foote et al., 2015; Yim et al., 2014). While the oxidative stress theory of aging posits that ROS generation in excess of antioxidant buffering capacity contributes to senescence and cell death, there are two directions from which this imbalance can originate (Finkel and Holbrook, 2000; Liochev, 2013). First, increased ROS generation may overwhelm existing antioxidant defenses, leading to oxidation of lipids, proteins and DNA. Second, antioxidant expression and activity may decline with advancing age, leaving organisms susceptible to oxidative damage. Age-related changes in specific antioxidants appear to be tissue, sex and species specific (Ji et al., 1990; Lawler and Demaree, 2001; reviewed in Sohal and Orr, 2012; Zhang et al., 2015), but aging and senescence are generally associated with a global decline in the capacity to respond to oxidative stress via Nrf2, a key regulator of the cellular antioxidant response (Suh et al., 2004; Zhang et al., 2015; Zhou et al., 2018a). In seals, such a decline would likely expose tissues to increased oxidative injury, as diving behavior does not decline in aging animals (Hindle et al., 2011). Several phocid species increase antioxidant expression throughout post-natal maturation, which coincides with the development of their physiological and behavioral dive capacities (Vázquez-Medina et al., 2010a, 2011a,b). These development-related increases in antioxidants may be an important component of seals' diving ability. While we selected only healthy animals – which could be expected to influence results as old age may contribute to outward signs of poor health – the inclusion of animals beyond peak reproductive age suggests that Weddell seals are generally able to maintain overall health beyond their reproductive peak. Relatedly, our previous work suggests that diving capacity does not decline with aging despite observed muscular senescence in these same animals (Hindle et al., 2009, 2011). Our results here show that old seals maintain antioxidant levels that may protect their tissues from the oxidative damage otherwise associated with aging. Consequently, seals may avoid functional losses and their fitness may not be compromised, even with advanced age.

Male and female adult Weddell seals demonstrated differential muscle-specific antioxidant strategies. In the longissismus dorsi, HSP70 abundance was elevated in males compared with levels in females. HSP70 is important in inhibiting apoptosis (Rérole et al., 2011), and in gray seals a greater abundance of HSP70 in tissues from suckling versus fasting pups appears to protect against protein oxidation (Bennett et al., 2014). Additionally, HSP70 is implicated in the skeletal muscle response to exercise stress (Puntschart et al., 1996; Senf, 2013). Male Weddell seals spend substantial time in the water guarding tidal cracks in the shorefast sea ice for access to breeding females (Siniff et al., 1977), potentially amounting to increased swimming times for males compared with females. Interestingly, no sex differences in apoptosis, 4-HNE-Lys levels or lipid hydroperoxide levels were detected in either longissimus dorsi or pectoralis muscles. Thus, increased HSP70 expression in the longissimus dorsi of male Weddell seals may reflect higher activity in this muscle and help limit associated cellular oxidative damage.

Glutathione (GSH) and its related enzymes are integral in the marine mammal response to redox changes. Murphy and Hochachka (1981) first noted higher GSH levels in the blood of Weddell seals compared with harbor seals (a shorter duration phocid diver) and humans. Since then, the glutathione antioxidant system has been extensively implicated in oxidative stress resistance in phocids (Vázquez-Medina et al., 2007, 2010b, 2011b,c). Genomic analyses have identified genes related to GSH metabolism under positive selection in marine mammals from independent lineages (Foote et al., 2015; Zhou et al., 2018b), and recent comparative work shows higher circulating GSH levels in marine compared with semi-aquatic and terrestrial mammals (García-Castañeda et al., 2017). In our study, GPx activity, which reduces lipid hydroperoxides (Birben et al., 2012; Brigelius-Flohé and Maiorino, 2013), was increased in the longissimus dorsi of males compared with females. Female Weddell seals displayed higher GR activity in longissimus dorsi compared with males. GR reduces oxidized glutathione (GSSG), promoting GSH recycling (Couto et al., 2016), suggesting that increased GSH recycling may help limit oxidative stress. In elephant seals, higher circulating lipid peroxidation levels have been found in males than in females sampled during the breeding season, accompanied by sex-specific increases in antioxidant activities across the season, during which time males engage in considerable fighting (Sharick et al., 2015). Here, we show that glutathione-related enzymes contribute to sex-specific antioxidant strategies in Weddell seal skeletal muscles, mitigating sex differences in oxidative damage.

In summary, our results demonstrate that the primary swimming muscle of old Weddell seals does not show increased oxidative stress despite observed apoptotic cell death and morphological senescence. Our results also show that males and females exhibit differential antioxidant profiles and that parameters related to redox metabolism are muscle type specific in seals. These results suggest that seals have evolved efficient and effective strategies to cope with diving- and age-associated increases in ROS generation and oxidative stress.

We thank the staff of the Albert P. Crary Science and Engineering Center at McMurdo Station for their support and assistance. R. Garrott and K. Proffitt shared data regarding age and lactation status for animals. We also thank two anonymous reviewers for their comments and feedback on the manuscript.