Skip to main content
NCBI home page
The Journal of Physiology logoLink to The Journal of Physiology
. 2018 Aug 28;596(24):6157–6171. doi: 10.1113/JP276539

Hepatic mitochondrial adaptations to physical activity: impact of sexual dimorphism, PGC1α and BNIP3‐mediated mitophagy

Alex Von Schulze 1,2, Colin S McCoin 1,2, Chiemela Onyekere 1, Julie Allen 1,2, Paige Geiger 1, Gerald W Dorn II 3, E Matthew Morris 1,2, John P Thyfault 1,2,✉
PMCID: PMC6292817  PMID: 30062822

Abstract

Key points

  • Hepatic mitochondrial adaptations to physical activity may be regulated by mitochondrial biogenesis (PGC1α) and mitophagy (BNIP3). Additionally, these adaptations may be sex‐dependent.

  • Chronic increase in physical activity lowers basal mitochondrial respiratory capacity in mice.

  • Female mice have higher hepatic electron transport system protein content, elevated respiratory capacity, lowered mitophagic flux, and emit less mitochondrial H2O2 independent of physical activity.

  • Males require chronic daily physical activity to attain a similar mitochondrial phenotype compared to females. In contrast, females have limited hepatic adaptations to chronic physical activity.

  • Livers deficient in PGC1α and BNIP3 display similar mitochondrial adaptations to physical activity to those found in wild‐type mice.

Abstract

Hepatic mitochondrial adaptations to physical activity may be regulated by biogenesis‐ and mitophagy‐associated pathways in a sex‐dependent manner. Here, we tested if mice with targeted deficiencies in liver‐specific peroxisome proliferator‐activated receptor γ coactivator 1α (PGC1α; LPGC1α+/−) and BCL2/adenovirus E1B 19 kDa protein‐interacting protein 3 (BNIP3)‐mediated mitophagy (BNIP3−/−) would have reduced physical activity‐induced adaptations in respiratory capacity, H2O2 emission and mitophagy compared to wild‐type (WT) controls and if these effects were impacted by sex. Male and female WT, LPGC1α+/− and BNIP3−/− C57BL6/J mice were divided into groups that remained sedentary or had access to daily physical activity via voluntary wheel running (VWR) (n = 6–10/group) for 4 weeks. Mice had ad libitum access to low‐fat diet and water. VWR reduced basal mitochondrial respiration, increased mitochondrial coupling and altered ubiquitin‐mediated mitophagy in a sex‐specific manner in WT mice. Female mice of all genotypes displayed higher electron transport system content, displayed increased ADP‐stimulated respiration, produced less mitochondrially derived reactive oxygen species, exhibited reduced mitophagic flux, and were less responsive to VWR compared to males. Males responded more robustly to VWR‐induced changes in hepatic mitochondrial function resulting in a match to adaptations found in females. Deficiencies in PGC1α and BNIP3 alone did not largely alter mitochondrial adaptations to VWR. However, VWR restored sex‐dependent abnormalities in mitophagic flux in LPGC1α+/−. Finally, BNIP3−/− mice had elevated mitochondrial content and increased mitochondrial respiration putatively through repressed mitophagic flux. In conclusion, hepatic mitochondrial adaptations to physical activity are more dependent on sex than PGC1α and BNIP3.

Keywords: metabolism, liver, steatosis, exercise, reactive oxygen species, mitochondrial respiratory capacity, female

Key points

  • Hepatic mitochondrial adaptations to physical activity may be regulated by mitochondrial biogenesis (PGC1α) and mitophagy (BNIP3). Additionally, these adaptations may be sex‐dependent.

  • Chronic increase in physical activity lowers basal mitochondrial respiratory capacity in mice.

  • Female mice have higher hepatic electron transport system protein content, elevated respiratory capacity, lowered mitophagic flux, and emit less mitochondrial H2O2 independent of physical activity.

  • Males require chronic daily physical activity to attain a similar mitochondrial phenotype compared to females. In contrast, females have limited hepatic adaptations to chronic physical activity.

  • Livers deficient in PGC1α and BNIP3 display similar mitochondrial adaptations to physical activity to those found in wild‐type mice.

Introduction

Physical activity improves hepatic mitochondrial function and prevents liver‐specific metabolic dysfunction (i.e. steatosis and insulin resistance) (Rector & Thyfault, 2011; Borengasser et al. 2012; Linden et al. 2015). However, the direct hepatic mechanisms associated with these beneficial adaptations remain unclear. It is thought that physical activity improves hepatic mitochondrial metabolism in part through the activation of mitochondrial biogenesis via increases in peroxisome proliferator‐activated receptor γ coactivator 1α (PGC1α) (Laye et al. 2009; Fletcher et al. 2014; Santos‐Alves et al. 2015). Specifically, induction of PGC1α in the liver results in the co‐activation of transcription factors known to target genes enhancing mitochondrial biogenesis, gluconeogenesis, fat oxidation and the electron transport system (ETS) (Fernandez‐Marcos & Auwerx, 2011). Indeed, we have shown that adenoviral overexpression of PGC1α leads to reduced lipid accumulation, increased hepatic mitochondrial fatty acid oxidation (FAO) and maximal respiratory capacity in hepatocytes (Morris et al. 2013), while others have shown that deficiencies in PGC1α lead to reduced hepatic mitochondrial content, FAO and defective gluconeogenesis (Leone et al. 2005; Burgess et al. 2006; Estall et al. 2009; Fletcher et al. 2018). Due to PGC1α’s association with both physical activity and hepatic metabolic health, it is critical to understand whether robust induction of PGC1α is required for the improved hepatic mitochondrial function observed with physical activity.

Physical activity may also improve hepatic mitochondrial function by enhancing organelle quality control through mitophagy, a subtype of macro‐autophagy characterized as the targeted degradation of damaged or low‐functioning mitochondria (Kasperek et al. 1982; He et al. 2012). It is theorized that mitochondrial biogenesis (via PGC1α) and mitochondrial degradation via mitophagy are tightly correlated (Kubli & Gustafsson, 2012). Specifically, it is thought that changes in mitochondrial biogenesis (via PGC1α) can lead to commensurate alterations in mitophagy (Vainshtein et al. 2015a), which may ultimately function as a rheostat for maintaining a healthy mitochondrial pool. One important mitophagic pathway is through the receptor‐mediated BCL‐2/adenovirus EIB 19 kDa interacting protein 3 (BNIP3). The importance of this pathway is highlighted by the increased hepatic oxidative stress, lipid accumulation and inflammation with corresponding reductions in FAO and respiration found in BNIP3−/− mice (Glick et al. 2012). However, it remains unclear if BNIP3 is activated in the liver by physical activity and is required for the induction of improved hepatic mitochondrial function.

Using wild‐type (WT) mice and mice with targeted deficiencies in liver‐specific PGC1α (LPGC1α+/−) and BNIP3‐mediated mitophagy (BNIP3−/−) we aimed to determine if intact mitochondrial biogenesis and BNIP3‐mediated mitophagy are required for the beneficial mitochondrial adaptations associated with physical activity in the liver. As previous evidence suggests that female mice may be protected from diet‐induced metabolic dysfunction in the liver (Hart‐Unger et al. 2017; Reue, 2017), we also chose to test if males or females displayed differential hepatic mitochondrial adaptations to physical activity in each genotype. An illustrated representation of the experimental purpose and design is shown in Fig. 1. We hypothesized that physical activity via voluntary wheel running (VWR) would increase maximal mitochondrial respiratory capacity and mitophagy in WT mice. Furthermore, we hypothesized that mice with deficiencies in liver PGC1α and BNIP3 would display reduced physical activity‐induced changes in maximal mitochondrial respiratory capacity as well as alterations in ETS protein content, mitophagy and mitochondrial reactive oxygen species (ROS) production levels compared to WT counterparts. Finally, because females are known to display greater lipid oxidation during exercise (Hedrington & Davis, 2015) and are known to be protected from steatosis (Hart‐Unger et al. 2017; Reue, 2017) we hypothesized that female mice would display greater physical activity‐induced changes in maximal mitochondrial respiratory capacity, ETS protein content and mitophagy compared to male counterparts.

Figure 1. Illustrated representation of experimental purpose and design.

Figure 1

Open in a new tab

This study examined the effects of increased physical activity through voluntary wheel running (VWR) on mitochondrial adaptations in male and female wild‐type (WT) mice, as well as mice with targeted deficiencies in liver specific PGC1α (LPGC1α+/−) and BNIP3‐mediated mitophagy (BNIP3−/−).

Methods

Ethical approval

The animal protocol was approved by the Institutional Animal Care and Use Committee at the University of Kansas Medical Center and Kansas City VA Medical Center (animal protocol number 2015‐2271). All experiments were carried out in accordance with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH guide, 8th edn, 2011), as well as The Journal of Physiology’s principles and standards for reporting animal experiments checklist. Mice were anaesthetized with pentobarbital sodium (100 mg kg−1) before a terminal procedure.

Animals

The liver‐specific, PGC1α heterozygous and BNIP3 null mice were developed as previously described (Diwan et al. 2007; Estall et al. 2009). Briefly, male WT C57Bl/6NJ mice were bred with female PGC1α homozygote floxed mice on a C57Bl/6NJ background (B6N.129(FVB)‐Ppargc1atm2.1Brsp/J; The Jackson Laboratory, Bar Harbor, ME, USA) resulting in female PGC1α heterozygote floxed mice, which were then bred with male albumin reporter‐driven Cre (B6.Cg‐Tg(Alb‐cre)21Mgn/J) to produce pups with liver‐specific heterozygosity for PGC1α. We elected to use a physiologically relevant heterozygote PGC1α knockdown because it is a transcriptional coactivator and it has been previously shown that this approach results in reduced expression of key genes controlling mitochondrial biogenesis and mitochondrial function, while maintaining appropriate glucose homeostasis (Estall et al. 2009; Fletcher et al. 2018). In contrast we elected to use a complete BNIP3 knockout because it is a protein directly involved in the process of mitophagy. PGC1α+/+ littermates of the LPGC1α+/− breeding paradigm were used as WT controls for all genotypes. Eight‐week‐old male and female WT, LPGC1α+/− and BNIP3−/− mice (n = 6–10 per sex, per strain) were bred and maintained on a C57Bl/6J background. Mice were housed at thermoneutrality (∼30°C) on a reverse light cycle (dark 10.00–22.00 h) with ad libitum access to water and low‐fat diet (LFD; D12110704: 10% kcal fat, 3.5% kcal sucrose and 3.85 kcal g−1 energy density, Research Diets, New Brunswick, NJ, USA). Animal weights, body composition and food intake were measured weekly throughout the 4‐week study. Body composition was measured at the beginning and end of the study by magnetic resonance imaging (model 900, EchoMRI, Houston, TX, USA). VWR mice were allowed free access to running wheels with distances recorded (ENV‐047 V, Med Associates Inc., Fairfax, VT, USA) and all mice were fasted for the last 4 h of the light cycle (06.00–10.00 h) before euthanasia and tissue collection.

Tissue collection and mitochondrial isolation

Livers of anaesthetized mice were quickly excised and either flash frozen in liquid N2 or placed into 8 mL ice cold mitochondrial isolation buffer (220 mM mannitol, 70 mM sucrose, 10 mM Tris, 1 mM EDTA, pH adjusted to 7.4 with KOH) and minced. Hepatic mitochondria were prepared as described previously (Morris et al. 2012). Briefly, livers were transferred to a 15 mL glass tube on ice and homogenized with a Teflon pestle. Homogenates were transferred to a 50 mL conical tube and centrifuged (4°C, 10 min, 1500 g). A portion of the supernatant was reserved (500 μL) in a fresh microtube and designated whole homogenate and the remaining supernatant was transferred into a 30 mL round‐bottom tube and centrifuged (4°C, 10 min, 8000 g). The pellet was resuspended in 5 mL isolation buffer using a glass‐on‐glass Dounce homogenizer for three to four passes and centrifuged (4°C, 10 min, 6000 g). The pellet was again resuspended in 4 mL isolation buffer containing 0.1% fatty acid‐free BSA using Dounce homogenization and centrifuged (4°C, 10 min, 4000 g). The final isolated mitochondrial pellet was resuspended in 500–750 μL modified MiR05 mitochondrial respiration buffer (0.5 mM EGTA, 3 mM MgCl2, 60 mM KMES, 20 mM glucose, 10 mM KH2PO4, 20 mM HEPES, 110 mM sucrose, 0.1% BSA, adjust pH to 7.1 with KOH). Protein concentrations were determined by bicinchoninic acid assay.

Mitochondrial respiration

Mitochondrial oxygen consumption (pmol s−1 mL−1) and H2O2 (pmol s−1 ml−1) flux were simultaneously measured using the Oroboros O2k fluorometer (Oroboros Instruments, Innsbruck, Austria) as described previously (Krumschnabel et al. 2015) and analysed using DatLab 7 (Oroboros Instruments). After calibration, mitochondrial isolates and appropriate substrates were added to the Oroboros chamber. All mitochondrial respiration experiments were completed with a chamber temperature of 30°C in the modified MiR05 mitochondrial respiration buffer described above with the addition of malate (2 mM), free CoA (63.5 μM), and l‐carnitine (2.5 mM) at a total volume of 2 mL. This modified MiR05 mitochondrial respiration buffer was utilized without the addition of lactobionic acid or taurine because they have antioxidant capacity and could possibly interfere with our H2O2 emission measurements. Coupled maximal respiration rate (state 3) was determined after the addition of adenosine 5′‐disphosphate (ADP; 2.5 mM). Maximal respiration rates for complex I were measured in two independent protocols utilizing either l‐palmitoylcarnitine (10 μM) or pyruvate (5 mM) in mitochondrial isolations. Maximal respiration through complex I and II (state 3S) was determined with the addition of succinate (10 mM). Maximal uncoupled respiration was determined with titrations of carbonyl cyanide‐p‐trifluoromethoxyphenylhydrazone (1 μM). All data were normalized to total or mitochondrial protein content within each chamber. The coupling control ratio, basal/state 3, was calculated to infer uncoupling or dyscoupling as previously described (Pesta & Gnaiger, 2012). H2O2:O2 ratios were calculated as state 3 respiration/state 3 H2O2 emission to infer the quantity of H2O2 emission at a given respiratory rate.

Determination of mitophagic flux

A separate cohort of mice undergoing the same paradigm described above was used for hepatic mitophagic flux analysis. Mitophagic flux was determined using in vivo intraperitoneal injections of the protease inhibitor leupeptin (LEU) as previously described (Haspel et al. 2011; Klionsky et al. 2016). Briefly, mice from all groups were treated with either 40 mg kg−1 LEU or an equal volume of the vehicle (saline) 18 h prior to sacrifice. An additional dose of 20 mg kg−1 LEU or equal volume of the vehicle was given 4 h prior to sacrifice to ensure continued autophagy inhibition. Mitophagic flux was inferred by the accumulation of mitophagic adapter proteins, microtubule associated protein 1 light chain 3B (LC3II) and sequestosome 1 (p62), in isolated mitochondrial samples (method described above) via western blotting.

Lipid peroxidation

Lipid peroxidation was determined by measuring malondialdehyde (MDA) end products. Briefly, a thiobarbituric acid MDA colorimetric reaction was used to determine total MDA levels in homogenized liver samples following the manufacturer's protocol (MAK085, Sigma‐Aldrich, St Louis, MO, USA).

Citrate synthase

Citrate synthase activity was determined in hepatic whole homogenate as described previously (Srere, 1969).

Liver triglyceride analysis

Hepatic triacylglycerol concentration was determined using a commercially available kit (TR0100, Sigma‐Aldrich), as described previously (Rector et al. 2008).

Western blotting

Tissue homogenization and mitochondrial isolation were completed as described above. Liver tissue was homogenized and stored in the mitochondrial isolation buffer described above. Mitochondrial isolations were resuspended and stored in the mitochondrial respiration buffer described above. Western‐ready Laemmli samples were produced from liver tissue homogenate and isolated mitochondria. SDS‐PAGE was used to separate samples, which were then transferred to polyvinylidene difluoride membrane and probed with primary antibodies at a concentration of 1:2,000. BNIP3 (3769S), parkin (4211S), mitofusin‐2 (MFN2; 9482S), dynamin‐related protein 1 (DRP1; 8570S), LC3A/B (12741S) and sequestosome 1 (SQSTM1)/p62 (5114S) antibodies were purchased from Cell Signaling Technology (Danvers, MA, USA). 4‐Hydroxynonenal (4‐HNE; ab46545) and total OXPHOS (MS604‐300) antibodies were purchased from Abcam (Cambridge, UK). Densitometry was used to quantify individual protein bands with Image Lab software (Bio‐Rad Laboratories, Hercules, CA, USA). All values were normalized to total protein or mitochondrial protein using 0.1% amido‐black (Sigma‐Aldrich) staining as previously described (Morris et al. 2012).

mRNA expression

RNA was extracted using an RNeasy mini‐kit following the manufacturer's instructions (74104, Qiagen, Hilden, Germany) and cDNA was prepared as previously described (Morris et al. 2013). A QuantStudio 3 Real‐Time PCR System (Thermo Fisher Scientific, Waltham, MA, USA) and SYBR green mouse primers (Sigma Aldrich) (Table 1) were used for real‐time quantitative PCR analysis. All mRNA values were normalized to the housekeeping gene, cyclophilin B.

Table 1.

SYBR RT‐PCR primer list

Gene Forward primer Reverse primer
MFN2 GTCATACCACCAATTGCTTC TCACAGTCTTGACACTCTTC
DNM1L GCGAACCTTAGAATCTGTGGACC CAGGCACAAATAAAGCAGGACGG
TFEB ACTATGATGGGGAAGAACAG GGTACTTGTACCTCCTTCTC
CAT CTCCATCAGGTTTCTTTCTTG CAACAGGCAAGTTTTTGATG
SOD2 CCATTTTCTGGACAAACCTG GACCTTGCTCCTTATTGAAG
GPX1 GGAGAATGGCAAGAATGAAG TTCGCACTTCTCAAACAATG
PPARGC1A TCACCATATTCCAGGTCAAG TCATAGGCTTCATAGCTGTC
PRKN GAGAAGAGCAGTACACTAGG CATGGTATGCTTCCTTACAG
MAP1LC3A AGTTGGTCAAGATCATCCG TCATCCTTCTCCTGTTCATAG
BNIP3 ACCACAAGATACCAACAGAG AATCTTCCTCAGACAGAGTG
SQSTM1 AATGTGATCTGTGATGGTTG GAGAGAAGCTATCAGAGAGG
Open in a new tab

Statistics

The main effects of sex, physical activity (VWR) and genotype were examined using three‐way ANOVA (SPSS Statistics, IBM Corp., Armonk, NY, USA). Upon discovering significant main effects, post hoc analyses were performed using least significant difference to test for any specific pairwise differences. Statistical significance was set at P < 0.05.

Results

Animal characteristics

Body mass (∼30%) and fat mass (∼82%) were higher in male mice (P < 0.05, Table 2). Male mice also displayed greater retroperitoneal fat mass, energy intake, serum cholesterol, serum triglycerides (TGs) and serum non‐esterified fatty acids (NEFAs) (P < 0.05, Table 2). As expected, VWR attenuated increases in body mass, total fat mass and retroperitoneal fat mass, while also increasing energy intake and heart mass for both sexes in all genotypes (P < 0.05, Table 2). Average daily running distance in VWR mice was similar between all groups (Table 2). The only genotype effect observed in anthropometric parameters was an attenuation of weight gain during the course of the study in LPGCα (reduced by ∼29% compared to WT, P < 0.05, Table 2).

Table 2.

Animal characteristics for wildtype (WT), LPGC‐1α+/− (LP), and BNIP3−/− (BN) following 4 weeks in sedentary (SED) or voluntary wheel running (VWR) conditions

Female Male
WT LP BN WT LP BN
Characteristic SED VWR SED VWR SED VWR SED VWR SED VWR SED VWR
Final body mass (g) 20.78 ± 0.82 21.71 ± 0.34 22.57 ± 0.76 21.08 ± 0.43 21.33 ± 0.57 21.86 ± 0.53 28.41 ± 1.79S,s 27.13 ± 0.89S,s 30.08 ± 1.85S,s 27.17 ± 0.80S,s 27.76 ± 0.72S,s 27.77 ± 0.59S,s
Weekly change body mass (g week−1) 0.37 ± 0.08 0.22 ± 0.06V 0.27 ± 0.08G 0.22 ± 0.07 V,G 0.23 ± 0.08 0.16 ± 0.09V 0.80 ± 0.22 0.20 ± 0.21V 0.41 ± 0.22G —0.30 ± 0.13 V,G,s,v 0.67 ± 0.12s 0.23 ± 0.11 V,v
Final fat mass (g) 2.07 ± 0.37 1.75 ± 0.14V 2.82 ± 0.60 1.70 ± 0.22V 1.62 ± 0.16 1.72 ± 0.15V 4.71 ± 1.29S 2.35 ± 0.58S,V 5.86 ± 1.27S 2.72 ± 0.48S,V,v 3.70 ± 0.30S,s 2.08 ± 0.25S,V,v
Weekly change fat mass (g week−1) 0.06 ± 0.06 −0.05 ± .06V 0.12 ± 0.05 −0.08 ± 0.07 V,v 0.04 ± 0.03 −0.03 ± 0.06V 0.44 ± 0.13s −0.23 ± 0.09 V,v 0.30 ± 0.16 −0.32 ± 0.16 V,v 0.31 ± 0.07s −0.02 ± 0.04 V,v
Heart mass/body mass (mg g−1) 4.25 ± 0.30 4.39 ± 0.10V 4.17 ± 0.41 4.25 ± 0.10V 4.40 ± 0.18s 4.42 ± 0.09V 4.13 ± 0.40 4.69 ± 0.19V 3.60 ± 0.22 4.38 ± 0.07 V,v 3.93 ± 0.10 4.40 ± 0.07 V,v
RP mass/body mass (mg g−1) 7.72 ± 0.54 5.75 ± 0.40 V,v 8.74 ± 1.22 5.52 ± 0.65 V,v 6.27 ± 0.53 5.65 ± 0.56V 9.26 ± 2.47S 6.03 ± 1.10S,V 12.33 ± 2.04S 6.59 ± 1.09S,V,v 8.87 ± 0.74S,s 5.51 ± 0.68S,V,v
Average weekly energy intake (kcal week−1) 43.68 ± 6.47 65.98 ± 5.82 V,v 46.53 ± 5.23 60.95 ± 6.39V 47.54 ± 6.84 62.74 ± 5.82V 61.43 ± 5.45S 81.77 ± 6.09S,V 55.87 ± 6.10S 67.22 ± 6.10S,V,v 56.71 ± 5.58S 78.32 ± 6.91S,V,v
Average daily running distance (km day−1) — 10.48 ± 1.50 — 10.08 ± 0.47 — 8.23 ± 0.34 — 9.07 ± 0.69 — 8.92 ± 0.82 — 8.90 ± 0.83
Serum cholesterol (mg dl−1) 63 ± 6 74 ± 4 77 ± 6 63 ± 7 52 ± 8 75 ± 7 111 ± 15S,s 98± 7S,s 96 ± 20S 93 ± 9S,s 122 ± 8S,s 105 ± 7S,s
Serum TGs (mg dl−1) 37 ± 2 34 ± 2 37 ± 4 38 ± 4 37 ± 2 34 ± 2 41 ± 4S 41 ± 2S,s 41 ± 4S 41 ± 3S 45 ± 2S,s 44 ± 3S,s
NEFA (mmol L−1) 0.32 ± 0.09 0.17 ± 0.02 0.18 ± 0.03 0.26 ± 0.04 0.16 ± 0.02 0.28 ± 0.05v,g 0.36 ± 0.08S 0.28 ± 0.05S 0.25 ± 0.03S 0.24 ± 0.04S 0.28 ± 0.03S,s 0.33 ± 0.04S
Open in a new tab

Data are presented as mean ± SEM (n = 6–10). S, P < 0.05 main effect of sex; V, P < 0.05 main effect VWR; G, P < 0.05 main effect of genotype; s, P < 0.05 sex within genotype within activity condition; g, P < 0.05 genotype within sex within activity condition; v, P < 0.05 VWR within sex within genotype. NEFA, non‐esterified fatty acid; TG, triglyceride; RP, retroperitoneal.

Female and BNIP3‐deficient mice have elevated hepatic electron transport system content

As we sought to characterize the impact of physical activity on mitochondrial adaptations in mice deficient in PGC1α and BNIP3 for both sexes, we first determined protein content of ETS proteins and citrate synthase enzyme activity in liver whole homogenate samples. Overall, females displayed markedly higher hepatic complex I–V protein content compared to males across all genotypes (main effect of sex, P < 0.05, Fig. 2 A–E). Additionally, male and female mice responded differently to VWR for complex V content (sex × VWR interaction, P < 0.05). BNIP3−/− mice displayed increased levels of complex I–V compared to WT (main effect of genotype, ∼15–30% higher, P < 0.05, Fig. 2 A–E). VWR caused a ∼29% and ∼27% reduction in hepatic protein content for complexes I and V in the LPGC1α+/− male mice, respectively (P < 0.05, Fig. 2 A and E). As inferred by citrate synthase activity, it appeared that overall mitochondrial content was similar between sexes (Fig. 2 F). However, male WT and LPGC1α+/− VWR mice had elevated citrate synthase activity when compared to their female counterparts (P < 0.05, Fig. 2 F). Citrate synthase activity was higher in BNIP3−/− mice for both sexes (P < 0.05, Fig. 2 F), suggesting increased mitochondrial content secondary to a deficiency in BNIP3‐mediated mitophagy.

Figure 2. Female mice have greater electron transport system content when compared to male mice.

Figure 2

Open in a new tab

Liver ETS content was determined for WT, LPGC1α+/− (LP) and BNIP3−/− (BN) mice using western immunoblotting for complex I (A), complex II (B), complex III (C), complex IV (D) and complex V (E). Western immunoblots (G) were normalized to total protein. Mitochondrial content was determined by evaluating citrate synthase activity (F). Data are presented as mean ± SEM (n = 6–10). S×V, P < 0.05 sex by VWR interaction; S×G, P < 0.05 sex by genotype interaction; V×G, VWR by genotype interaction; S, P < 0.05 main effect of sex; V, P < 0.05 main effect VWR; G, P < 0.05 main effect of genotype; s, P < 0.05 sex within genotype within activity condition; g, P < 0.05 genotype within sex within activity condition; v, P < 0.05 VWR within sex within genotype.

Female mice have enhanced maximal respiratory capacity while PGC1α and BNIP3 deficiencies do not impact adaptations

To support our findings in ETS content and citrate synthase activity, we determined whether maximal mitochondrial respiratory capacity was similarly dichotomous between sexes and if there was a differential response to physical activity between genotypes. Importantly, we found that maximal basal respiratory capacity was actually decreased by VWR (main effect of VWR, P < 0.05, Figs. 3 A and 4 A). Furthermore, females displayed higher basal, state 3 and state 3S maximal respiratory capacity with palmitoylcarnitine (PCarn) compared to males (P < 0.05, Fig. 3 A–C). Females also displayed higher basal, state 3, state 3S and uncoupled maximal respiratory capacity with pyruvate when compared to males (P < 0.05, Fig. 4 D). Interestingly, female mice were less responsive to VWR in terms of coupling efficiency. In contrast, VWR in the males tended to enhance coupling efficiency for both PCarn and pyruvate substrates. Overall, physical activity‐induced adaptations for coupling control ratios in the males resulted in values that were similar to females in both sedentary (SED) and VWR groups (Figs. 3 I and 4 I). The primary genotype differences observed were that deficiencies in BNIP3 led to elevated state 3S maximal respiratory capacity with pyruvate compared to the WT (main effect of genotype, P < 0.05, Fig. 4 C) and deficiencies in PGC1α resulted in elevated state 3 oxygen flux with PCarn (P < 0.05, Fig. 3 B).

Figure 3. Female mice have elevated maximal respiratory capacity and reduced H2O2 emission compared to male mice using PCarn.

Figure 3

Open in a new tab

Liver maximal mitochondrial respiratory capacity was determined in WT, LPGC1α+/− (LP) and BNIP3−/− (BN) mice using the Oroboros O2k fluorometer in isolated mitochondrial samples using PCarn as the substrate. Basal (A), ADP‐stimulated state 3 (B), ADP+succinate‐stimulated state 3S (C) and uncoupled (D) maximal respiratory capacity were measured. State 2 (no ADP) (E), ADP‐stimulated state 3 (F), ADP+succinate‐stimulated state 3S (G) and uncoupled (H) H2O2 emission were also measured. Coupling control ratios (I) and state 3 H2O2:O2 flux ratios (J) were also calculated. Maximal respiratory capacity and H2O2 emission values were normalized to the mitochondrial protein content added to each chamber as measured via the BCA assay. Data are presented as mean ± SEM (n = 6–10). S×V, P < 0.05 sex by VWR interaction; S×G, P < 0.05 sex by genotype interaction; V×G, VWR by genotype interaction; S, P < 0.05 main effect of sex; V, P < 0.05 main effect VWR; G, P < 0.05 main effect of genotype; s, P < 0.05 sex within genotype within activity condition; g, P < 0.05 genotype within sex within activity condition; v, P < 0.05 VWR within sex within genotype.

Figure 4. Female mice have elevated maximal respiratory capacity and reduced H2O2 emission compared to male mice using pyruvate.

Figure 4

Open in a new tab

Liver maximal mitochondrial respiratory capacity was determined in WT, LPGC1α+/− (LP) and BNIP3−/− (BN) mice using the Oroboros O2k fluorometer in isolated mitochondrial samples using pyruvate as the substrate. Basal (A), ADP‐stimulated state 3 (B), ADP+succinate‐stimulated state 3S (C) and uncoupled (D) maximal respiratory capacity were measured. State 2 (no ADP) (E), ADP‐stimulated state 3 (F), ADP+succinate‐stimulated state 3S (G) and uncoupled (H) H2O2 emission were also measured. Coupling control ratios (I) and state 3 H2O2:O2 flux ratios (J) were also calculated. Maximal respiratory capacity and H2O2 emission values were normalized to the mitochondrial protein content added to each chamber as measured via the BCA assay. Data are presented as mean ± SEM (n = 6–10). S×V, P < 0.05 sex by VWR interaction; S×G, P < 0.05 sex by genotype interaction; V×G, VWR by genotype interaction; S, P < 0.05 main effect of sex; V, P < 0.05 main effect VWR; G, P < 0.05 main effect of genotype; s, P < 0.05 sex within genotype within activity condition; g, P < 0.05 genotype within sex within activity condition; v, P < 0.05 VWR within sex within genotype.

Female mice have reduced mitochondrial H2O2 emission regardless of genotype

To provide further clarity surrounding coupling efficiency and oxidative stress, we measured real‐time mitochondrial H2O2 emission in both our PCarn and pyruvate substrate protocols. We observed marked reductions in basal, state 3, state 3S and uncoupled H2O2 emission in the female mitochondrial samples compared to males in all genotypes for both PCarn and pyruvate substrates (main effect of sex, P < 0.05, Figs. 3 E–H and 4E–H). Importantly when compared to males, female mice emit significantly less H2O2 while having higher maximal respiratory capacity as shown by a ratio of H2O2 emission: O2 consumption (Figs. 3 J and 4 J). Finally, there was a main effect of VWR across all genotypes when comparing H2O2 emission: O2 consumption in the pyruvate protocol (P < 0.05, Fig. 4 J). Counter to our hypothesis that deficiency in BNIP3‐mediated mitophagy would increase mitochondrial ROS production, we saw no difference in H2O2 emission for the BNIP3−/− mice (Figs. 3 E–H and 4E–H). Interestingly, we also found that state 3 H2O2 emission: O2 consumption was reduced in SED mice deficient in PGC1α with PCarn substrate (P < 0.05, Fig. 3 J). Importantly, deficiencies in PGC1α and BNIP3 did not alter VWR‐induced changes to H2O2 emission, but again these effects were driven by a lowering in females (Fig. 3 E–H and 4E–H).

Sex differences in mitochondrial H2O2 emission are likely driven by improved mitochondrial coupling efficiency

To delineate possible causes of the variation in H2O2 emission between sexes, we examined mRNA expression of genes encoding enzymes critical for mitigating H2O2 (SOD2, CAT and GPX1). Interestingly, SOD2 mRNA was not different between males and females (Table 3). However, CAT mRNA was higher in males while GPX1 was higher in females (∼25% and ∼19% increase, respectively, P < 0.05, Table 3). Importantly, GPX1 was significantly lower in the male WT and LPGC1α+/− VWR mice. Thus, it is possible that a VWR‐induced reduction in GPX1 contributed to increased H2O2 emission: O2 consumption in the pyruvate protocol. It is also possible that VWR‐induced depletion of glutathione contributed to increased H2O2 emission: O2 consumption for the VWR mice in the pyruvate protocol (Lew et al. 1985); however, the adaptive increase in glutathione due to chronic training should prevent this (Lima et al. 2013). Combined, these data suggest that the increased coupling efficiency observed in the females is likely the primary driver of their reduced H2O2 emission rather than significant alteration in antioxidant enzyme profiles. Importantly, antioxidant gene expression was not dramatically altered with deficiencies in PGC1α or BNIP3 (Table 3).

Table 3.

mRNA expression of liver autophagy/mitophagy, oxidative stress and mitochondrial biogenesis genes

Female Male
WT LP BN WT LP BN
Gene SED VWR SED VWR SED VWR SED VWR SED VWR SED VWR
MFN2 0.92 ± 0.17 1.03 ± 0.11 1.08 ± 0.09 1.04 ± 0.14 0.94 ± 0.12 1.45 ± 0.26 1.28 ± 0.20 1.18 ± 0.18 1.21 ± 0.24 0.94 ± 0.09 1.15 ± 0.21 0.97 ± 0.11
DNM1L 0.97 ± 0.20 1.00 ± 0.12 1.17 ± 0.16 0.99 ± 0.12 0.92 ± 0.12G 1.02 ± 0.10G 1.20 ± 0.13 1.01 ± 0.15 1.03 ± 0.14 0.84 ± 0.15 0.70 ± 0.08G,g 0.74 ± 0.09G
TFEB 1.10 ± 0.21S 1.07 ± 0.12S 1.26 ± 0.17S 0.99 ± 0.12S 1.03 ± 0.05S,s 1.23 ± 0.17S 1.07 ± 0.16 0.82 ± 0.09 1.03 ± 0.18 0.85 ± 0.09 0.78 ± 0.05 0.88 ± 0.12
CAT 0.73 ± 0.08 0.88 ± 0.08 0.92 ± 0.11 0.87 ± 0.08 0.78 ± 0.07G 0.79 ± 0.06G 1.27 ± 0.15S,s 1.19 ± 0.15S 1.28 ± 0.19S 0.95 ± 0.10S 0.75 ± 0.10S,G,g 0.81 ± 0.10S,G
SOD2 0.99 ± 0.14 0.92 ± 0.06 1.07 ± 0.09 0.95 ± 0.08 0.95 ± 0.09G 0.95 ± 0.06G 1.09 ± 0.08 1.03 ± 0.11 1.12 ± 0.13 0.79 ± 0.07v 0.79 ± 0.07G,g 0.77 ± 0.07G
GPX1 1.05 ± 0.15S 1.09 ± 0.12S,s 1.14 ± 0.14S 1.03 ± 0.12S,s 0.95 ± 0.13S 1.03 ± 0.10S 1.05 ± 0.10 0.77 ± 0.06v 1.15 ± 0.16 0.71 ± 0.09v 0.77 ± 0.08 0.85 ± 0.09
PPARGC1A 1.30 ± 0.28S 1.29 ± 0.19S 0.82 ± 0.14S 1.00 ± 0.15S 1.09 ± 0.22S,s 1.45 ± 0.32S,s 0.99 ± 0.15 1.09 ± 0.28 1.09 ± 0.24 0.68 ± 0.22 0.52 ± 0.08g 0.43 ± 0.08g
PRKN 1.06 ± 0.19 1.49 ± 0.20 1.36 ± 0.23 1.30 ± 0.16 0.94 ± 0.18G 0.99 ± 0.10G 1.05 ± 0.06 1.32 ± 0.30 1.16 ± 0.19 1.11 ± 0.17 0.64 ± 0.08G,g 0.74 ± 0.10G
MAP1LC3A 1.03 ± 0.11 0.86 ± 0.11 1.19 ± 0.14 0.88 ± 0.09 0.93 ± 0.09 1.16 ± 0.15 1.12 ± 0.17 0.98 ± 0.19 0.96 ± 0.14 0.77 ± 0.12 1.11 ± 0.17 0.88 ± 0.09
BNIP3 1.01 ± 0.16 1.03 ± 0.14 1.65 ± 0.40 1.23 ± 0.20 0.00 ± 0.00G,g 0.00 ± 0.00G,g 1.22 ± 0.26 0.97 ± 0.27 1.71 ± 0.62 0.77 ± 0.21 0.00 ± 0.00G,g 0.00 ± 0.00G,g
SQSTM1 1.12 ± 0.31 1.03 ± 0.19 1.54 ± 0.33 1.11 ± 0.21 1.37 ± 0.34 0.94 ± 0.20 1.59 ± 0.50 1.23 ± 0.36 1.83 ± 0.67 1.15 ± 0.35 0.93 ± 0.19 0.94 ± 0.25
Open in a new tab

Gene expression was determined using RT‐PCR. Relative expression of listed liver autophagy/mitophagy, oxidative stress and mitochondrial biogenesis genes were normalized to cycophilin B (PPIB) and are presented as mean ± SEM (n = 6–10). S, P < 0.05 main effect of sex; V, P < 0.05 main effect VWR; G, P < 0.05 main effect of genotype; s, P < 0.05 sex within genotype within activity condition; g, P < 0.05 genotype within sex within activity condition; v, P < 0.05 VWR within sex within genotype.

Male mice require VWR to reduce mitophagic flux to levels similar to female mice regardless of genotype

As the process of mitophagy is closely related to mitochondrial function and health, we assessed whether sex or physical activity via VWR modulated mitophagic flux and mitophagic protein/mRNA content within our genotypes. As autophagy/mitophagy is a transient process, the use of an autophagic inhibitor is required to properly determine autophagic flux (Klionsky et al. 2016). In accordance with previous literature (Haspel et al. 2011), we elected to use the protease inhibitor LEU to block macroautophagy. Additionally, to capture mitophagy we measured the mitophagic adapter proteins LC3II and p62 in mitochondrial isolates as opposed to whole homogenate samples. We first confirmed the efficiency of leupeptin (LEU) to appropriately block lysosomal degradation (Fig. 5C). We then found that the mitophagy adapter proteins LC3II and p62 were both significantly higher in male mice treated with LEU (main effect of sex, P < 0.05, Fig. 5 B and E). While not significant, it did appear that VWR reduced p62 levels in male mice treated with LEU (Fig. 7 E). In non‐LEU treated mice, LC3II and p62 content were no different or higher, likely due to reduced mitophagy‐mediated degradation (Fig. 5 A and D).

Figure 5. Male mice have increased mitophagy flux compared to females, which is attenuated by VWR.

Figure 5

Open in a new tab

To measure autophagy flux, LC3II and p62 were measured in the absence (A and D, respectively) and presence (B and E, respectively) of the autophagy inhibitor leupeptin (LEU). Western immunoblots (C) were normalized to total protein or mitochondrial protein dependent on the sample fraction analysed. Data are presented as mean ± SEM (n = 6–10). S×V, P < 0.05 sex by VWR interaction; S×G, P < 0.05 sex by genotype interaction; V×G, VWR by genotype interaction; S, P < 0.05 main effect of sex; V, P < 0.05 main effect VWR; G, P < 0.05 main effect of genotype; s, P < 0.05 sex within genotype within activity condition; g, P < 0.05 genotype within sex within activity condition; v, P < 0.05 VWR within sex within genotype.

Figure 7. Male mice require VWR to reduce liver triglycerides when compared to females.

Figure 7

Open in a new tab

Liver lipid content was determined for WT, LPGC1α+/− (LP) and BNIP3−/− (BN) mice using haematoxylin and eosin stained slides (A), and biochemical liver TG assays (B). Liver oxidative stress was determined using malondialdehyde (MDA) assays (C) and 4‐hydroxynonenal (4‐HNE) immunoblotting (D and E). Immunoblots were normalized to total protein. Data are presented as mean ± SEM (n = 6–10). S×V, P < 0.05 sex by VWR interaction; S×G, P < 0.05 sex by genotype interaction; V×G, VWR by genotype interaction; S, P < 0.05 main effect of sex; V, P < 0.05 main effect VWR; G, P < 0.05 main effect of genotype; s, P < 0.05 sex within genotype within activity condition; g, P < 0.05 genotype within sex within activity condition; v, P < 0.05 VWR within sex within genotype.

Consistent with reduced mitophagy flux, accumulation of BNIP3 and parkin protein was higher in female mice, which we believe indicates a lack of mitophagy‐mediated degradation (∼47% and ∼75% higher, respectively, P < 0.05, Fig. 6 A and B). Parkin accumulation was potentiated with VWR in WT female mice, but this effect was not seen in the other genotypes (Fig. 6 B). Additionally, female mice had higher levels of MFN2 and DRP1, which was potentiated by VWR (main effect of sex, P < 0.05, Fig. 6 D and E). Hepatic mRNA expression of MFN2, DNM1L, PRKN, MAP1LC3A, BNIP3 and SQSTM1 was not significantly different between sexes and was not altered by VWR (Table 3).

Figure 6. Female mice have reduced mitophagy‐mediated degradation of BNIP3 and parkin compared to male mice.

Figure 6

Open in a new tab

Liver mitophagy was determined in WT, LPGC1α+/− (LP) and BNIP3−/− (BN) using western immunoblotting for mitophagy‐associated proteins in whole homogenate and isolated mitochondrial samples. BNIP3 (A), parkin (B), MFN2 (D) and DRP1 (E) were measured in whole homogenate samples. Western immunoblots (C) were normalized to total protein or mitochondrial protein dependent on the sample fraction analysed. Data are presented as mean ± SEM (n = 6–10). S×V, P < 0.05 sex by VWR interaction; S×G, P < 0.05 sex by genotype interaction; V×G, VWR by genotype interaction; S, P < 0.05 main effect of sex; V, P < 0.05 main effect VWR; G, P < 0.05 main effect of genotype; s, P < 0.05 sex within genotype within activity condition; g, P < 0.05 genotype within sex within activity condition; v, P < 0.05 VWR within sex within genotype.

Deficiencies in PGC1α and BNIP3 significantly alter mitophagic flux

With regards to the impact of deficiencies in PGC1α and BNIP3, we found that LC3II content was higher in SED males and females deficient in PGC1α treated with and without LEU, respectively (P < 0.05, Fig. 5 A and B). This effect was normalized to values similar to the WT with VWR. Interestingly, there was a dramatic reduction in p62 content in the SED LPGC1α+/− females, which again was normalized to levels similar to the WT with VWR (Fig. 6 E). Deficiencies in PGC1α resulted in the accumulation of BNIP3 and parkin, which counter to the WT animals, we believe indicates an increased rate of mitophagic flux (due to increased flux rates seen with LEU) (∼48% and ∼52% higher, respectively, P < 0.05, Fig. 6 A and B). Importantly, VWR consistently reduced BNIP3 content in the LPGC1α+/− mice, agreeing with the LEU injection/flux data that VWR reduces mitophagic flux in the PGC1α‐deficient male mice (P < 0.05, Fig. 6 A). Collectively, it appeared that a critical adaptation to VWR in PGC1α‐deficient animals restores mitophagic equilibrium to rates similar to the WT. Additionally, this compensatory adaptation appears to occur via different mechanisms between male and female mice.

Mice deficient in BNIP3 did not have altered LC3II flux, but did have reduced p62 flux compared to the WT (Fig. 6 B and E). In line with this, BNIP3−/− mice also displayed reduced parkin protein levels (female VWR only, P < 0.05, Fig. 6 B) and PRKN transcript levels (main effect of genotype P < 0.05, Table 3) suggesting that ubiquitin‐mediated mitophagy via parkin/p62 is reduced). Deficiencies in BNIP3 also resulted in a reduction in DRP1 content with a commensurate increase MFN2, possibly causing mitochondrial hyperfusion (P < 0.05, Fig. 6 D and E). Collectively, it appears that deficiencies in BNIP3‐mediated mitophagy result in a compensatory reduction in alternative mitophagy pathways to maintain mitochondrial content. Additionally, hyperfusion of mitochondria in the BNIP3‐deficient animals may prevent the global mitochondrial dysfunction that we expected to occur.

Sex, VWR and deficiencies in PGC1α and BNIP3 contribute to liver oxidative stress and liver triglycerides

Finally, to determine whole‐liver effects of sex, VWR, and deficiencies in PGC1α and BNIP3, we measured hepatic triglycerides (TGs) and markers of oxidative stress. VWR reduced liver TG ∼50% in all males (P < 0.05, Fig. 7 B), but WT and BNIP3−/− females were unresponsive to VWR. Liver TGs were also higher in SED LPGC1α+/− mice compared to SED WT mice for both sexes; however, VWR reduced liver TGs (P < 0.05, Fig. 7 B). There was no effect of BNIP3−/− on liver TGs compared to WT.

Counter to our mitochondrial data, we found that hepatic MDA and 4‐HNE levels were higher in female mice across all genotypes (P < 0.05, Fig. 7 C and D). LPGC1α+/− mice also displayed higher levels of 4‐HNE than WT mice for both sexes (P < 0.05, Fig. 7 D). VWR increased 4‐HNE content in female WT mice and MDA content in BNIP3−/− females (∼11% and 12% higher in VWR, respectively, P < 0.05, Fig. 7 C and D). However, VWR had no effect on markers of oxidative stress in male mice.

Discussion

This study aimed to determine whether intact mitochondrial biogenesis (PGC1α) and BNIP3‐mediated mitophagy are required for physical activity‐induced mitochondrial adaptions in the liver. Additionally, we examined if hepatic mitochondrial adaptations to physical activity are sex‐dependent. Here we show that female mice have higher hepatic ETS protein content, have elevated maximal respiratory capacity, have altered mitophagic flux, and emit less mitochondrial H2O2 independent of physical activity and genotype. Importantly, we also show that males require physical activity to attain a similar mitochondrial phenotype compared to females in all genotypes. Finally, we determined that deficiencies in PGC1α and BNIP3 largely do not alter the sex‐dependent mitochondrial response to physical activity. However, mice deficient in PGC1α require physical activity to restore sex‐dependent abnormalities in mitophagic flux and mice deficient in BNIP3 have increased mitochondrial content putatively through repressed mitochondrial turnover.

Both a classic study and a previous report from our group have shown that physical activity increases hepatic maximal mitochondrial respiratory capacity in rats (Glick, 1966; Fletcher et al. 2014). However, here we show that physical activity does not increase respiratory capacity in mice. In fact, we observed that basal respiratory capacity was reduced with physical activity and that this occurred in both PGC1α‐ and BNIP3‐deficient mice. A potential explanation for the discrepancy is that mice have higher rates of energy expenditure, a limited ability to store muscle glycogen and a greater reliance on hepatic gluconeogenesis to maintain glucose homeostasis (Kowalski & Bruce, 2014). Thus, mitochondrial respiratory capacity likely already operates at a near maximal level that cannot be further increased with exercise. Of greater importance, we show here that there are fundamental differences in the hepatic mitochondrial adaptation to physical activity between male and female mice.

Previous literature showed that female mice are more resistant to diet‐induced hepatic steatosis (Hart‐Unger et al. 2017; Reue, 2017), suggesting that females may possess differences in hepatic metabolism (mitochondrial content, respiratory capacity or turnover). In addition, the demands of gestation and lactation have likely led to evolutionary differences in the capacity for hepatic mitochondria to adapt differently from those of males (Maggi & Della Torre, 2018). Indeed, we found that female mice had higher hepatic ETS protein content, increased maximal mitochondrial respiratory capacity using both PCarn and pyruvate substrates, and marked reductions in mitochondrial H2O2 emission. Importantly, we show that males require VWR to obtain mitochondrial coupling efficiencies similar to females. It is possible that 4 weeks of physical activity reduces electron leak and decreases the degree of oxidative metabolism required to maintain mitochondrial membrane potential in males, while females maintain tighter coupling control independent of physical activity. Thus, the primary adaptation to physical activity in male mice may be improved mitochondrial quality rather than increased respiratory capacity.

One putative way to improve mitochondrial quality is the targeted degradation of damaged mitochondria via mitophagy. Recent reports highlight that mitophagy is critical for exercise‐induced mitochondrial adaptations in skeletal muscle (Fu et al. 2018). However, little attention has been placed on the liver despite higher rates of mitochondrial turnover (Miwa et al. 2008). One recent study investigated hepatic autophagy/mitophagy in response to VWR (Rosa‐Caldwell et al. 2017), but only males were examined and, critically, an autophagy inhibitor to assess mitophagic flux was not utilized. Here, we examined the effects of physical activity on a variety of mitophagy‐associated proteins in both sexes for all genotypes. We also directly assessed mitophagy flux by measuring the mitophagy‐associated adapter proteins LC3II and p62 in the presence and absence of the autophagy inhibitor LEU. Critically, we found that mitophagy is sex‐dependent.

With the use of injections of LEU as an inhibitor to block lysosomal degradation, we are the first to show that female mice, regardless of SED and VWR conditions, have evidence of reduced mitochondrial turnover. As with mitochondrial coupling efficiency, males require physical activity to have the reduced ubiquitin‐mediated (p62) mitophagy flux that is seen in females in the SED condition. These data are supported by the observation that female livers have higher accumulation of mitophagy proteins (BNIP3 and parkin) compared to male mice. Moreover, increased MFN2 and DRP1 content in the females suggests that they may rely more on fission and fusion rather than mitophagy to maintain mitochondrial health. Combined, these data suggest that chronic physical activity leads to increased coupling efficiency and less need for mitochondrial turnover in male mice, but that physical activity does not impact these factors in female mice. Importantly, these trends remain true in mice with deficiencies in PGC1α and BNIP3. Our results highlight the importance of using an autophagy inhibitor to measure LC3II and p62, as well as measuring multiple mitophagy‐associated proteins, to provide a clear idea of mitophagic flux.

As mitochondrial biogenesis and mitochondrial degradation via autophagy/mitophagy are thought to aid mitochondrial health, and this process is modulated by exercise in skeletal muscle (Kubli & Gustafsson, 2012; Vainshtein et al. 2015a, b ), we tested if targeted deficiencies in PGC1α and BNIP3 impacted the sex‐dependent mitochondrial adaptations to physical activity. We found that deficiencies in PGC1α and BNIP3 largely do not impact the sex‐specific mitochondrial adaptations to physical activity in mice. However, similar to other groups we found that deficiencies in liver PGC1α led to elevated liver TGs in SED conditions (Leone et al. 2005; Burgess et al. 2006; Estall et al. 2009; Fletcher et al. 2018). Importantly, we show that physical activity prevents hepatic TG accumulation in LPGC1α+/− mice despite null changes in maximal respiratory capacity. Despite the known transcriptional coordination between PGC1α and transcription factor EB to induce autophagy (Settembre et al. 2013), we also found that deficiencies in liver PGC1α actually enhanced mitophagic flux in SED LPGC1α+/− males and reduced ubiquitin‐mediated mitophagic flux in SED LPGC1α+/− females. As physical activity reversed these mitophagy‐related effects, it is possible that physical activity restores mitophagic equilibrium and plays a role in reducing hepatic lipid content. Alternatively, other known benefits of physical activity may have contributed to the lower TG levels including reduced hepatic de novo lipogenesis, increased lipid utilization by skeletal muscle, and improved regulation of lipolysis during postprandial conditions (Rector & Thyfault, 2011). Our evidence that physical activity‐induced mitochondrial adaptations can still occur in animals with deficiencies in liver PGC1α have important clinical implications as data suggests that obese human subjects chronically possess reduced hepatic PGC1α levels (Croce et al. 2007).

To directly evaluate the role of mitophagy in the hepatic response to physical activity, we tested a model deficient in BNIP3‐mediated mitophagy (BNIP3−/− mice). As mentioned above, the loss of BNIP3 did not largely alter sex‐specific mitochondrial adaptations to physical activity. However, in agreement with others (Glick et al. 2012), we found that BNIP3−/− mice had increased mitochondrial content, putatively due to downregulated mitophagy, and increased maximal respiratory capacity (with PCarn substrate). Unlike Glick et al. we did not observe steatosis after a 4 h fast and observed no changes in H2O2 emission or tissue ROS levels (Glick et al. 2012). Again, physical activity caused a reduction in liver TGs for the BNIP3−/− males compared to their sedentary counterparts with minimal changes in mitochondrial respiratory capacity. Like the LPGC1α+/− mice, this suggests that physical activity reduces hepatic TG content independent of changes in measured hepatic respiratory capacity. Interestingly, markers of mitochondrial biogenesis (PPARGC1A mRNA) and ubiquitin‐mediated mitophagic flux were blunted in the BNIP3−/− males, which we believe is a compensatory adaptation to increased mitochondrial content. BNIP3−/− mice also displayed altered fusion and fission profiles as determined by increased MFN2 and reduced DRP1 content. This adaptation may have allowed the BNIP3−/− mice to maintain mitochondrial respiratory capacity through enhanced fusion despite being deficient in BNIP3‐mediated mitophagy.

Conclusion

The primary findings of the current study are that physical activity causes a reduction in basal mitochondrial respiratory capacity (both sexes), and that female mice have higher hepatic ETS protein content, have elevated maximal respiratory capacity, have altered mitophagic flux, and emit less mitochondrial H2O2 independent of physical activity and genotype. Importantly, we also show that males require physical activity to attain a similar mitochondrial phenotype compared to females in WT mice and those with deficiencies in PGC1α and BNIP3. Livers deficient in PGC1α and BNIP3 largely have similar mitochondrial to physical activity compared to WT. However, we found that deficiencies in PGC1α led to sex‐dependent abnormalities in mitochondrial turnover, which can be restored to equilibrium (WT levels) by physical activity. Finally, mice with deficiencies in BNIP3‐mediated mitophagy have increased mitochondrial content putatively through repressed mitochondrial turnover. In summary, mitochondrial adaptations to physical activity are more dependent on sex than PGC1α‐ and BNIP3‐mediated processes. Further research is needed to determine if these sex‐dependent effects play a role in the known effect of female rodents of being protected from hepatic steatosis (Hart‐Unger et al. 2017; Reue, 2017), and if these sex‐dependent effects are due to oestrogen signaling or some other factors unique to female mice. Progressive research in this area may provide the data necessary to develop novel therapeutic compounds to prevent metabolic complications in the liver.

Additional information

Competing interests

The authors have no conflict of interest related to this manuscript.

Author contributions

This work was all performed in the laboratory of J.P.T. which is affiliated with the Kansas City VA Medical Center and is located at the University of Kansas Medical Center. Conception and design of the work A.V.S., C.S.M., G.W.D., E.M.M. and J.P.T.; acquisition, analysis, or interpretation A.V.S., C.S.M., C.O., J.A., P.G., E.M.M. and J.P.T.; the manuscript was drafted and revised critically by A.V.S., C.S.M., C.O., J.A., P.G., G.W.D., E.M.M. and J.P.T. All authors approved the final version of the manuscript. All authors agree to be accountable for all aspects of the work and ensure that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All persons designated as authors qualify for authorship and all those who qualify for authorship are listed.

Funding

This work was supported by VA Merit Review grant (1I01BX002567‐01) (J.P.T.) and also partially supported by NIH grants (R01DK088940) (J.P.T.), (K01DK112967‐01) (E.M.M.), Institutional Development Award (IDeA) (NIGMS P20 GM103418) (J.P.T., E.M.M. and C.S.M.) and CTSA TL2 postdoctoral training grant (TL1TR002368) (C.S.M.), and (NIH R35 HL135736) (G.W.D.). Additional support was provided by the KUMC Biomedical Research Training Program (C.S.M.) and University of Kansas Self Fellowship program (A.V.S.).

Biographies

Alex Von Schulze is pursuing a PhD in Molecular and Integrative Physiology at the University of Kansas Medical Center under the mentorship of J.P.T. and P.G. He received a BS in Biomedical science from Northern Arizona University and a MS in Health and Sports Science from the University of Memphis. He is interested in understanding the molecular mechanisms by which physical activity and mitochondrial health can prevent metabolic dysfunction in humans.

graphic file with name TJP-596-6157-g001.gif

Colin S. McCoin earned his PhD in Molecular, Cellular and Integrative Physiology at UC Davis examining the role of intermediate metabolite acylcarnitines in eliciting inflammation and insulin resistance in skeletal muscle. Currently, he is a postdoctoral fellow studying the effects of exercise and obesity on hepatic and skeletal muscle mitochondrial function under the mentorship of J.P.T. at the University of Kansas Medical Center.

Edited by: Michael Hogan & Bettina Mittendorfer

A. Von Schulze and Colin S. McCoin are co‐first authors.

This is an Editor's Choice article from the 15 December 2018 issue.

Linked articles This article is highlighted by a Perspectives article by Bellissimo & Perry. To read this Perspective, visit https://doi.org/10.1113/JP276896.

References

  1. Borengasser SJ, Rector RS, Uptergrove GM, Morris EM, Perfield JW 2nd, Booth FW, Fritsche KL, Ibdah JA & Thyfault JP (2012). Exercise and omega‐3 polyunsaturated fatty acid supplementation for the treatment of hepatic steatosis in hyperphagic OLETF rats. J Nutr Metab 2012, 268680. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Burgess SC, Leone TC, Wende AR, Croce MA, Chen Z, Sherry AD, Malloy CR & Finck BN (2006). Diminished hepatic gluconeogenesis via defects in tricarboxylic acid cycle flux in peroxisome proliferator‐activated receptor gamma coactivator‐1α (PGC‐1α)‐deficient mice. J Biol Chem 281, 19000–19008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Croce MA, Eagon JC, LaRiviere LL, Korenblat KM, Klein S & Finck BN (2007). Hepatic lipin 1β expression is diminished in insulin‐resistant obese subjects and is reactivated by marked weight loss. Diabetes 56, 2395–2399. [DOI] [PubMed] [Google Scholar]
  4. Diwan A, Krenz M, Syed FM, Wansapura J, Ren X, Koesters AG, Li H, Kirshenbaum LA, Hahn HS, Robbins J, Jones WK & Dorn GW (2007). Inhibition of ischemic cardiomyocyte apoptosis through targeted ablation of Bnip3 restrains postinfarction remodeling in mice. J Clin Invest 117, 2825–2833. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Estall JL, Kahn M, Cooper MP, Fisher FM, Wu MK, Laznik D, Qu L, Cohen DE, Shulman GI & Spiegelman BM (2009). Sensitivity of lipid metabolism and insulin signaling to genetic alterations in hepatic peroxisome proliferator‐activated receptor‐gamma coactivator‐1α expression. Diabetes 58, 1499–1508. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Fernandez‐Marcos PJ & Auwerx J (2011). Regulation of PGC‐1α, a nodal regulator of mitochondrial biogenesis. Am J Clin Nutr 93, 884s–890s. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Fletcher JA, Linden MA, Sheldon RD, Meers GM, Morris EM, Butterfield A, Perfield JW 2nd, Rector RS & Thyfault JP (2018). Fibroblast growth factor 21 increases hepatic oxidative capacity but not physical activity or energy expenditure in hepatic peroxisome proliferator‐activated receptor gamma coactivator‐1α‐deficient mice. Exp Physiol 103, 408–418. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Fletcher JA, Meers GM, Linden MA, Kearney ML, Morris EM, Thyfault JP & Rector RS (2014). Impact of various exercise modalities on hepatic mitochondrial function. Med Sci Sports Exerc 46, 1089–1097. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Fu T, Xu Z, Liu L, Guo Q, Wu H, Liang X, Zhou D, Xiao L, Liu L, Liu Y, Zhu MS, Chen Q & Gan Z (2018). Mitophagy directs muscle‐adipose crosstalk to alleviate dietary obesity. Cell Rep 23, 1357–1372. [DOI] [PubMed] [Google Scholar]
  10. Glick D, Zhang W, Beaton M, Marsboom G, Gruber M, Simon MC, Hart J, Dorn GW 2nd, Brady MJ & Macleod KF (2012). BNip3 regulates mitochondrial function and lipid metabolism in the liver. Mol Cell Biol 32, 2570–2584. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Glick JL (1966). Effects of exercise on oxidative activities in rat liver mitochondria. Am J Physiol 210, 1215–1221. [DOI] [PubMed] [Google Scholar]
  12. Hart‐Unger S, Arao Y, Hamilton KJ, Lierz SL, Malarkey DE, Hewitt SC, Freemark M & Korach KS (2017). Hormone signaling and fatty liver in females: analysis of estrogen receptor α mutant mice. Int J Obes (Lond) 41, 945–954. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Haspel J, Shaik RS, Ifedigbo E, Nakahira K, Dolinay T, Englert JA & Choi AM (2011). Characterization of macroautophagic flux in vivo using a leupeptin‐based assay. Autophagy 7, 629–642. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. He C, Bassik MC, Moresi V, Sun K, Wei Y, Zou Z, An Z, Loh J, Fisher J, Sun Q, Korsmeyer S, Packer M, May HI, Hill JA, Virgin HW, Gilpin C, Xiao G, Bassel‐Duby R, Scherer PE & Levine B (2012). Exercise‐induced BCL2‐regulated autophagy is required for muscle glucose homeostasis. Nature 481, 511–515. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hedrington MS & Davis SN (2015). Sexual dimorphism in glucose and lipid metabolism during fasting, hypoglycemia, and exercise. Front Endocrinol 6, 61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Kasperek GJ, Dohm GL, Barakat HA, Strausbauch PH, Barnes DW & Snider RD (1982). The role of lysosomes in exercise‐induced hepatic protein loss. Biochem J 202, 281–288. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Klionsky DJ, Abdelmohsen K, Abe A, et al (2016). Guidelines for the use and interpretation of assays for monitoring autophagy (3rd edition). Autophagy 12, 1–222. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kowalski GM & Bruce CR (2014). The regulation of glucose metabolism: implications and considerations for the assessment of glucose homeostasis in rodents. Am J Physiol Endocrinol Metab 307, E859–E871. [DOI] [PubMed] [Google Scholar]
  19. Krumschnabel G, Fontana‐Ayoub M, Sumbalova Z, Heidler J, Gauper K, Fasching M & Gnaiger E (2015). Simultaneous high‐resolution measurement of mitochondrial respiration and hydrogen peroxide production. Methods Mol Biol 1264, 245–261. [DOI] [PubMed] [Google Scholar]
  20. Kubli DA & Gustafsson ÅB (2012). Mitochondria and mitophagy: The yin and yang of cell death control. Circ Res 111, 1208–1221. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Laye MJ, Rector RS, Borengasser SJ, Naples SP, Uptergrove GM, Ibdah JA, Booth FW & Thyfault JP (2009). Cessation of daily wheel running differentially alters fat oxidation capacity in liver, muscle, and adipose tissue. J Appl Physiol (1985) 106, 161–168. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Leone TC, Lehman JJ, Finck BN, Schaeffer PJ, Wende AR, Boudina S, Courtois M, Wozniak DF, Sambandam N, Bernal‐Mizrachi C, Chen Z, Holloszy JO, Medeiros DM, Schmidt RE, Saffitz JE, Abel ED, Semenkovich CF & Kelly DP (2005). PGC‐1α deficiency causes multi‐system energy metabolic derangements: muscle dysfunction, abnormal weight control and hepatic steatosis. PLoS Biol 3, e101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Lew H, Pyke S & Quintanilha A (1985). Changes in the glutathione status of plasma, liver and muscle following exhaustive exercise in rats. FEBS Lett 185, 262–266. [DOI] [PubMed] [Google Scholar]
  24. Lima FD, Stamm DN, Della‐Pace ID, Dobrachinski F, de Carvalho NR, Royes LFF, Soares FA, Rocha JB, González‐Gallego J & Bresciani G (2013). Swimming training induces liver mitochondrial adaptations to oxidative stress in rats submitted to repeated exhaustive swimming bouts. PLoS One 8, e55668. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Linden MA, Fletcher JA, Morris EM, Meers GM, Laughlin MH, Booth FW, Sowers JR, Ibdah JA, Thyfault JP & Rector RS (2015). Treating NAFLD in OLETF rats with vigorous‐intensity interval exercise training. Med Sci Sports Exerc 47, 556–567. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Maggi A & Della Torre S (2018). Sex, metabolism and health. Mol Metab 15, 3–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Miwa S, Lawless C & von Zglinicki T (2008). Mitochondrial turnover in liver is fast in vivo and is accelerated by dietary restriction: application of a simple dynamic model. Aging Cell 7, 920–923. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Morris EM, Jackman MR, Meers GM, Johnson GC, Lopez JL, MacLean PS & Thyfault JP (2013). Reduced hepatic mitochondrial respiration following acute high‐fat diet is prevented by PGC‐1α overexpression. Am J Physiol Gastrointest Liver Physiol 305, G868–G880. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Morris EM, Meers GM, Booth FW, Fritsche KL, Hardin CD, Thyfault JP & Ibdah JA (2012). PGC‐1α overexpression results in increased hepatic fatty acid oxidation with reduced triacylglycerol accumulation and secretion. Am J Physiol Gastrointest Liver Physiol 303, G979–G992. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Pesta D & Gnaiger E (2012). High‐resolution respirometry: OXPHOS protocols for human cells and permeabilized fibers from small biopsies of human muscle. Methods Mol Biol 810, 25–58. [DOI] [PubMed] [Google Scholar]
  31. Rector RS & Thyfault JP (2011). Does physical inactivity cause nonalcoholic fatty liver disease? J Appl Physiol (1985) 111, 1828–1835. [DOI] [PubMed] [Google Scholar]
  32. Rector RS, Thyfault JP, Morris RT, Laye MJ, Borengasser SJ, Booth FW & Ibdah JA (2008). Daily exercise increases hepatic fatty acid oxidation and prevents steatosis in Otsuka Long‐Evans Tokushima Fatty rats. Am J Physiol Gastrointest Liver Physiol 294, G619–G626. [DOI] [PubMed] [Google Scholar]
  33. Reue K ( 2017). Sex differences in obesity: X chromosome dosage as a risk factor for increased food intake, adiposity and co‐morbidities. Physiol Behav 176, 174–182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Rosa‐Caldwell ME, Lee DE, Brown JL, Brown LA, Perry RA Jr, Greene ES, Carvallo Chaigneau FR, Washington TA & Greene NP (2017). Moderate physical activity promotes basal hepatic autophagy in diet‐induced obese mice. Appl Physiol Nutr Metab 42, 148–156. [DOI] [PubMed] [Google Scholar]
  35. Santos‐Alves E, Marques‐Aleixo I, Rizo‐Roca D, Torrella JR, Oliveira PJ, Magalhaes J & Ascensao A (2015). Exercise modulates liver cellular and mitochondrial proteins related to quality control signaling. Life Sci 135, 124–130. [DOI] [PubMed] [Google Scholar]
  36. Settembre C, De Cegli R, Mansueto G, Saha PK, Vetrini F, Visvikis O, Huynh T, Carissimo A, Palmer D, Klisch TJ, Wollenberg AC, Di Bernardo D, Chan L, Irazoqui JE & Ballabio A (2013). TFEB controls cellular lipid metabolism through a starvation‐induced autoregulatory loop. Nat Cell Biol 15, 647–658. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Srere PA ( 1969). [1] Citrate synthase: [EC 4.1.3.7. Citrate oxaloacetate‐lyase (CoA‐acetylating)]. Methods Enzymol 13, 3–11. [Google Scholar]
  38. Vainshtein A, Desjardins EM, Armani A, Sandri M & Hood DA (2015a). PGC‐1α modulates denervation‐induced mitophagy in skeletal muscle. Skelet Muscle 5, 9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Vainshtein A, Tryon LD, Pauly M & Hood DA (2015b). Role of PGC‐1α during acute exercise‐induced autophagy and mitophagy in skeletal muscle. Am J Physiol Cell Physiol 308, C710–C719. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society

RESOURCES