All institutional and governmental regulations regarding the ethical use of were complied with during this study, which was approved by the Institutional Animal Care and Use Committee of Yeungnam University (YUMC-AEC2018-034). Male ICR mice weighing ∼28 g and C57BL6 mice weighing ∼20 g were purchased form Koatech (Gyeonggi, South Korea), ICR mice were used to analyze lipid profile and histochemistry in adipose tissue, and C57BL6 mice were used to determine circulating glucose disposal and molecular phenotype in skeletal muscle. Mice were housed under a 12:12-h light/dark cycle and provided a chow diet and water ad libitum. GW501516 (GW) was administered in diet at 40 mg/kg, and 5% EtOH was administered ad libitum for 5 weeks.

C₂C₁₂ mouse cells were purchased from ATCC (cat. no. CRL-1772; Manassas, VA, United States), and cell study was performed as previously described (Koh et al., 2019a). Briefly, cells were maintained in 5% CO₂ at 37°C and grown in DMEM (Sigma-Aldrich, St. Louis, MO, United States) containing 10% fetal bovine serum (Sigma-Aldrich, St. Louis, MO, United States), penicillin (100 U/ml), and streptomycin (100 μg/ml) (Thermo Fisher Scientific, Waltham, MA, United States). C₂C₁₂ cells were differentiated into myotubes by changing the medium to 2% horse serum (Thermo Fisher, Waltham, MA, United States) containing penicillin (100 U/ml) and streptomycin (100 μg/ml) (Thermo Fisher, Waltham, MA, United States); 50 mM or 30 mM of EtOH was treated in myotubes.

We used pcDNA 3.1-Myc-PPARδ vector, which was constructed as previously described (Koh et al., 2019a). HyPer-mito (Evrogen, Moscow, Russia), which is a sensor to detect mitochondrial hydrogen peroxide (H₂O₂), is a fluorescent protein and was gifted by Dr. Dong-Hyung Cho at Kyungpook National University, South Korea. pcDNA 3.1-empty vector and Myc-PPARδ vector were transfected using Lipofectamine 2000 (Thermo Fisher, Waltham, MA, United States).

Plasma free fatty acid and triglyceride concentrations were measured using a kit obtained from Waco Chemicals (Richmond, VA, United States).

Adipose tissues were fixed with 10% formalin solution, embedded in paraffin blocks, and sectioned at 4 μm. Hematoxylin and eosin and Masson’s trichrome staining were performed using standard methods. Digital images were captured using an Aperio CS digital pathology slide scanner (Leica, Germany). Adipocyte size was analyzed using the ImageJ program (NIH, Bethesda, MD, United States).

After 5 weeks of EtOH with or without GW or saline treatment, mice were fasted overnight, and blood samples were collected from a tail vein to determine basal blood glucose levels. We performed IPGTT as previously described (Shin et al., 2019). Briefly, glucose (1.5 g/kg body weight) was injected intraperitoneally, and blood samples were collected at 15, 30, 60, 90, and 120 min. Blood glucose levels were measured using OneTouch blood glucose meters (LifeScan Europe, Switzerland).

Muscle triglyceride concentrations were determined by extracting total lipids from clamp-frozen muscle samples using chloroform-methanol (2:1 vol/vol), as described by Folch et al. (1957). After separating chloroform and methanol-water phases and phospholipid removal, samples were processed using Frayn and Maycock’s modification (Frayn and Maycock, 1980) of the Denton and Randle method (Denton and Randle, 1967). Triglycerides were quantified spectrophotometrically using an enzymatic assay kit (Waco Chemicals, Richmond, VA, United States).

Glucose uptake by myotubes was determined using ¹⁴C-2-deoxy-D-glucose (2DG, PerkinElmer Life and Analytical Sciences, United States). The glucose uptake method described below has been reproduced in part from the previous study (Shin et al., 2019). Briefly, myotubes were fasted for 2 h in serum-free medium, treated with insulin (100 nM) for 1 h, ¹⁴C-2DG was then added to each well, and 20 mM cytochalasin B was added to control cells. Cells were lysed with 60 mM NaOH and insulin-stimulated myotube glucose uptakes were determined using a Liquid Scintillation Analyzer (PerkinElmer).

Myotubes were fasted for 2 h in serum-free medium containing insulin (100 nM) for 1 h and then harvested to evaluate AKT activation.

Western blots were performed as previously described (Koh et al., 2019a). Briefly, frozen tissues were powdered and then homogenized at 15:1 (vol/wt) in ice-cold RIPA buffer supplemented with protease inhibitor cocktail (Thermo Fisher Scientific, Houston, TX, United States). Protein contents were measured using a DC-protein assay kit (Bio-Rad). The antibodies used were as follows: AMPK (cat. no. 2793), phospho-AMPK (pAMPK; cat. no. 2541), mammalian target of rapamycin (mTOR; cat. no. 2972), phosphor-mTOR (p-mTOR; cat. no. 2971), regulatory associated protein of mTOR (Raptor; cat no. 2280), phosphor-Ser792-Raptor (p-Raptor; cat. no. 2083), rapamycin-insensitive companion of mammalian target of rapamycin (Rictor; cat. no. 2114), phosphor-Thr1135-Rictor (p-Rictor; cat. no. 3806), phosphor-Thr308-Akt (pT308-Akt; cat. no. 13038), phosphor-Ser473-Akt (pS473-Akt; cat. no. 4060), and Akt (cat. no. 2920) from Cell Signaling Technologies (Danvers, MA, United States); all Cell Signaling antibody were diluted 1:1,000. Superoxide dismutase 2 (SOD2; cat. no. sc-137254. 1:500 dilution) and carnitine palmitoyltransferase (CPT1; cat. no. 393070, 1:1,000 dilution) were from Santa Cruz Biotechnology. β-actin (cat. no. A5441, 1:1,000 dilution) was from Sigma Aldrich (St. Louis, MO, United States). Long-chain acyl-CoA dehydrogenase (LCAD; cat. no. ab196655, 1:1,000 dilution) and glucose transporter type 4 (GLUT4; cat. no. ab-654, 1:1,000 dilution) were from Abcam (Cambridge, MA, United States). Citrate synthase (CS; cat. no. CISY11-A, 1:5,000 dilution) was from Alpha Diagnostics (San Antonio, TX, United States). ATP synthase (cat. no. 459240, 1:5,000 dilution), succinate ubiquinone oxidoreductase (SUO; cat. no. 459200, 1:5,000 dilution), and NADH ubiquinone oxidoreductase (NUO; cat. no. 459210, 1:5,000 dilution) were from Thermo Fisher Scientific (Waltham, MA, United States). Uncoupling protein 3 (UCP3; cat. no. 10750-AP, 1:1,000 dilution) and carnitine-acylcarnitine translocase (CACT; cat. no. 19363-AP, 1:1,000 dilution) were from Proteintech (Rosemont, IL, United States).

We performed hydrogen peroxide emission analysis as previously published (Koh et al., 2019b). Briefly, palmitate or saline treated PPARδ or empty vector (EV) myotubes in 75 t flasks were washed with DPBS and suspended in 0.25% trypsin-EDTA medium. Myotubes were then centrifuged, re-suspended in respiration medium [105 mM K-MES, 30 mM KCL, 10 mM KH2PO4, 5 mM MgCl2-6H2O, 5 mg/ml BSA] pH 7.4 supplemented with 1 mM EGTA (pH 7.3)] including 10 mM Amplex Red, horseradish peroxidase 1 U/ul, superoxide dismutase 10 U/ul, and treated with 500 μg/ml digitonin (∼50% TLC, Sigma, St. Louis, United States) to permeabilize membranes. Glutamate (5 mM, 2 mmol/L) and malate (1 mmol/L) were then added to stimulate H₂O₂ production under state 4 respiration conditions. A fluorolog 3 (Horiba Jobin Yvon, Edison, Piscataway, NJ, United States) spectrofluorometer was used to monitor Amplex Red (Invitrogen, Carlsbad, United States) oxidation in myotubes.

Fresh tissues were homogenized in first isolation buffer containing 10 mM EDTA, 215 mM D-mannitol, 75 mM sucrose, 20 mM HEPES, and 0.1% free-fatty acid BSA (pH 7.4) using Potter-Elvehjem tissue homogenizer and then centrifuged at 700 g for 10 min at 4°C followed by transfer supernatant to new microcentrifuge tube. The supernatant was centrifuged at 10,500 g for 10 min, then supernatants were discarded and pellets (mitochondria) were resuspended in 100 μl of second isolation buffer [3 mM EDTA, 215 mM D-mannitol, 75 mM sucrose, 20 mM HEPES, and 0.1% free-fatty acid BSA (pH 7.4)].

Mitochondrial oxygen consumption was assessed using an Oxytherm System (Hansatech Instruments), as previously described (Finlin et al., 2018). Briefly, mitochondria were added to the Oxytherm chamber containing 500 μl of mitochondrial respiration buffer (125 mM KCl, 2 mM MgCl₂, 2.5 mM KH₂PO₄, 20 mM HEPES, pH adjusted to 7.2) at 37°C under constant stirring. Complex I-driven respiration was initiated by adding pyruvate/malate (5 mM and 2.5 mM, respectively) or palmitoyl-carnitine/malate (5 mM and 2.5 mM, respectively) followed by ADP (to 300 μM in 2 steps) to induce State 3 respiration. ATP synthase activity was then inhibited by adding 2.5 μM oligomycin to induce State 4 respiration. Finally, maximum respiration was measured in the presence of 10 μM trifluoromethoxy carbonylcyanide phenylhydrazone (FCCP). OCRs were determined and expressed as nanomoles/min. Mitochondrial respiratory uncoupling via UCP3 was calculated as the difference between State 3 respiration (with FFA as substrate) and State 4 respiration (induced by oligomycin).

To test the effects of chronic alcohol consumption, mice consumed 5% EtOH for 4 weeks. Food intake was lower in EtOH consumption mice than control (Figure 1A), but water consumption was higher (min–max; 138–152 per 4 weeks) in 5% EtOH consumption group than control (min–max; 174–195 per 4 weeks) (Figure 1B). EtOH treatment did not changes body weights (BW, Figure 1D) because no difference of total calorie intake (Figure 1C). EtOH consumption slightly decreases visceral fat weight (Figure 1E) but significantly increased plasma free fatty acid (FFA) levels as compared with controls (Figure 1H). Since the source the increased FFA observed in plasma was adipose tissue, we examined epididymal adipocyte sizes by hematoxylin-eosin staining. Representative sections showed that area of adipocytes in 5% EtOH consumption mice were significantly smaller than in controls (Figures 1F,G). We also found EtOH consumption increased the activation of hormone sensitive lipase in rats (Supplementary Figure S1). Furthermore, we observed that triglyceride (TG) levels in mouse skeletal muscle (Figure 1K) and liver (Figure 1J) were higher than in controls, but not plasma TG (Figure 1I). These results show 5% EtOH consumption induced the accumulation of TG in skeletal muscle and that this accumulation might be associated with elevated FFA level resulted from increased lipolysis in adipose tissue.

We next studied whether EtOH directly influences various metabolic enzymes and signaling factors related to fatty acid and glucose metabolism using C₂C₁₂ myotubes. Western blotting was used to assess AMPK, mTOR, CPT1, GLUT4, and SOD2 protein. It has been documented that AMPK, a cellular energy sensor, clearly link to muscle glucose uptake (Koh et al., 2019a) and metabolism (Gan et al., 2011) as well as fatty acid oxidation and mitochondrial biogenesis (O’Neill et al., 2013). We found that AMPK is significantly decreased by 50 mM EtOH in myotubes. Muscle-specific CPT1 knockout mice have been shown to exhibit low levels of fatty acid oxidation and increased lipid accumulation in the muscle (Wicks et al., 2015), these exceed fatty acid accumulation results in a decrease in GLUT4 content (Koh et al., 2019b), as expected, CPT1 and GLUT4 in myotubes were decreased by 50 mM EtOH treatment (Figure 2A). However, since the myotubes were incubated with high glucose, fatty acid may not be accumulated much in the myotubes by EtOH treatment, and thus, it appears that GLUT4 protein synthesis was inhibited by lack of mTOR signaling induced by 50 mM EtOH (Figure 2A), furthermore, ROS can also affect the protein degradation (Kocha et al., 1997), therefore, it is possible that 50 mM EtOH may decrease GLUT4 content via lack of mTOR signaling and/or ROS. Cellular energy metabolism is tightly regulated by substrate concentrations and signaling intensity, and thus, we studied the temporal effects of a lower concentration of EtOH (30 mM). We observed that this lower concentration slightly decreased the phosphorylation of AMPK. However, GLUT4, phosphorylated mTOR, Raptor, and Rictor amount in myotubes exposed to this lower concentration were significantly diminished after 6 h of treatment and then gradually recovered (Figures 2B–G), presumably because almost all of the ethanol appeared to have been metabolized at 48 h. These results suggest EtOH toxicity can directly downregulate signaling related to energy metabolism and protein synthesis even lower amount EtOH (Figure 2).

PPARδ activation has shown to increase fatty acid oxidation (Fan et al., 2017) and PPARδ overexpression has been reported to increase glucose uptake (Koh et al., 2019a). Since acute 5% EtOH consumption reduces metabolic regulation factors related to PPARδ including pAMPK, CPT1, and GLUT4 (Figure 2), we speculated that PPARδ can directly or indirectly involved in metabolic disorder by chronic alcohol consumption. We observed mice treated with 5% EtOH for 5 weeks had significantly higher TG levels in muscle tissue than controls, but GW plus 5% EtOH treated mice had lower TG levels than 5% EtOH treated mice (Figure 3A). IPGTT results showed the circulating glucose disposal rate in 5% EtOH treated mice was lower than in non-treated controls, but that co-treatment with GW maintained glucose disposal rates in 5% EtOH treated mice at the non-treated control level (Figures 3B,C). Since skeletal muscle responsible for approximately 80% of insulin-mediated glucose uptake in the postprandial state (Thiebaud et al., 1982), we tested insulin sensitivity in mouse muscle cells according to the treatment of EtOH and GW (Figures 3D–G). We found that glucose uptake induced by insulin in EtOH 30 mM treated myotubes was significantly lower than in saline treated myotubes (Figure 3H). However, PPARδ overexpression in myotubes significantly increased glucose uptake level in both of insulin dependent and independent manner (Figure 3H). To identify whether the mechanism responsible for this glucose uptake was regulated by EtOH or PPARδ, we examined AKT signaling induced by insulin. It was found 30 mM EtOH did not interrupt insulin-induced threonine 308 AKT (t308-AKT) signaling, and that insulin-induced increases in t308-AKT occurred in PPARδ overexpressing myotubes (Figures 3D,E). In contrast, serine 473 AKT (S473-AKT) phosphorylation was inhibited by EtOH, and this effect was reversed by PPARδ (Figures 3F,G). We found PPARδ accelerated the normalization of Rictor (Figure 3I) but not Raptor activation in EtOH exposed-myotubes (Figure 3J).

We also observed that EtOH did not decrease GLUT4 protein content after 24 h of treatment of EtOH same as Figure 2D; however, PPARδ overexpression in myotubes increased GLUT4 protein amount regardless of EtOH and saline treatment (Figure 3K). Together, these results suggest the effects of PPARδ including the restoration of insulin mediated-lower S473-AKT signaling and increase GLUT4 protein content may affect insulin-stimulated glucose disposal rate in blood.

We were curious that PPARδ activation can induce modification of the mitochondrial respiration chain to protect cellular against metabolic disorder induced by EtOH. We tested oxygen consumption rates in mitochondria extracted from the muscles of EtOH and/or GW treated mice, and found neither EtOH nor GW influenced mitochondrial respiration (Figures 4A–C) or respiration control ratio [RCR, indicator of the coupling state of mitochondria (Rhein et al., 2010)] (Figure 4D) when pyruvate and malate (PM) was used as a substrate. However, state 4 (Figure 4F) and maximal oxygen respiration (Figure 4G) of mitochondria from EtOH plus GW and from GW treated mice were higher than in non-treated controls when palmitoyl-carnitine and malate (PCM) was used as substrate. We also found coupling efficiency in mitochondria from GW treated mice was lower than in non-treated control mice when PCM was used as substrate (Figure 4H). Next, we sought to determine potential mechanisms by which GW influences mitochondrial respiration. Contrary to that observed in the cell study, AMPK activation and CACT protein content were not down-regulated by EtOH, presumably because we isolated muscles 36 h after alcohol treatment to eliminate the acute effects of alcohol. CACT transports acylcarnitine derived from fatty acid oxidation into the mitochondrial matrix (Ogawa et al., 2000), and the previous study has shown that CACT deficiency dysregulates fatty acid oxidation (Pande et al., 1993), in the present study, CACT protein content in mouse muscle was unaffected by EtOH treatment, but GW increased CACT protein amount when administered with water or EtOH (Figure 4I). LCAD is a key enzyme in mitochondrial β-oxidation (Lea et al., 2000), and we observed LCAD protein content were lower in the muscles of EtOH treated mice than in non-treated controls, but not in EtOH plus GW treated mice (Figure 4I). Since GW consumption is known to increase fatty acid oxidation (Fan et al., 2017), we did not expect this result, and thus, we investigated the effect of GW treatment with or without palmitate on AMPK activation and LCAD protein content in myotubes. We found that palmitate down-regulates AMPK phosphorylation and LCAD protein content, but GW increases AMPK activation regardless of palmitate presence, and LCAD content was recovered by GW with palmitate (Figure 4J). These results suggest GW activation is regulated in a fatty acid concentration-dependent manner, and thus, GW may not be activated much in the low or normal level of TG concentration in the tissue (Figures 3A, 4I). No differences in citrate synthase (CS; involved in the tricarboxylic acid (TCA) cycle), NU), or SUO protein content were observed in skeletal muscle (Figure 4I). ATPsyn amount was slightly lower in 5% EtOH treated mouse muscle than in non-treated controls, but identical in EtOH plus GW treated mice and non-treated controls (Figure 4I). Furthermore, GW significantly increased uncoupling protein 3 (UCP3) levels when treated alone or with EtOH (Figure 4I). Collectively, GW appears to induce mitochondrial respiration uncoupling via UCP3 and to enhance fatty acid oxidation in mitochondria when fatty acids are used as substrates (Figure 4).

Reactive oxygen species emissions from mitochondria can induce insulin resistance (Anderson et al., 2009). We determined H₂O₂ emission from myotubes and found palmitate increased H₂O₂ emission (Figure 5A), but that PPARδ blocked H₂O₂ emission induced by palmitate (Figure 5A). We also observed that PPARδ activation increased catalase protein content in skeletal muscle of both water and EtOH consumed mice (Figure 5B). To test whether PPARδ activation by GW can reduce H₂O₂, a plasmid of mitochondria H₂O₂ sensor (HyPer-mito) was transfected into C₂C₁₂ muscle cells. We found EtOH increased HyPer-mito fluorescence intensity and that this was recovered by GW (Figure 5C). These data indicate that PPARδ protects muscles against both palmitate and EtOH induced ROS emission via catalase.