Exercise and exercise training-induced increase in autophagy markers in human skeletal muscle

trained male subjects (mean age 25 years; range 19–33 years) com- pleted an 8-week exercise training intervention consisting of continuous mod- erate cycling at 157 20 W for 60 min (MOD;n=6) or continuous moderate cycling (157 20 W) interspersed by 30-sec sprints (473 79 W) every 10 min (SPRINT;n=6) 3 days per week. Sprints were followed by 3:24 min at 102 17 W to match the total work between protocols. A muscle biopsy was obtained before, immediately and 2 h after the first training session as well as at rest after the training session. In both MOD and SPRINT, skeletal muscle AMPK Thr172 and ULK Ser317 phosphorylation was elevated immediately after exercise, whereas mTOR Ser2448 and ULK Ser757 phosphorylation was unchanged. Two hours after exercise LC3I, LC3II and BNIP3 protein content was overall higher than before exercise with no change in p62 protein. In MOD, Beclin1 protein content was higher immediately and 2 h after exercise than before exercise, while there were no differences within SPRINT. Oxphos complex I, LC3I, BNIP3 and Parkin protein content was higher after the train- ing intervention than before in both groups, while there was no difference in LC3II and p62 protein. Beclin1 protein content was higher after the exercise training intervention only in MOD. Together this suggests that exercise increases markers of

exercise, while there were no differences within SPRINT. Oxphos complex I, LC3I, BNIP3 and Parkin protein content was higher after the train- ing intervention than before in both groups, while there was no difference in LC3II and p62 protein. Beclin1 protein content was higher after the exercise training intervention only in MOD. Together this suggests that exercise increases markers of autophagy in human skeletal muscle within the first 2 h of recovery and 8 weeks of exercise training increases the capacity for autop- hagy and mitophagy regulation. Hence, the present findings provide evidence that exercise and exercise training regulate autophagy in human skeletal muscle and that this in general was unaffected by interspersed sprint bouts. Introduction Endurance exercise training increases the oxidative capacity of skeletal muscle in part through increased mitochondrial biogenesis in skeletal muscle with con- comitant health beneficial effects and enhanced exercise performance (Booth et al. 2012). Thus, exercise training has been shown to increase the content and/or activity of oxidative proteins such as OXPHOS complexes, cyto- chrome (Cyt) c, and citrate synthase (CS) in human and rodent skeletal muscle (Hood 2001). Moreover, exercise training has been suggested to regulate mito- chondrial quality control mechanisms. This includes enhanced autophagy and mitophagy, which is the selective removal of damaged mitochondria (Drake et al. 2016). Autophagy is exerted by a complex machinery of multi- ple components. The autophagosome formation is an important step in autophagy and the binding of UNC51- like kinase (ULK)1 protein and the Class III PI 3-kinase (PI3K) complex, including Beclin1, has been shown to be involved in formation of autophagosomes (Nakahira and Choi 2013). AMP-activated protein kinase (AMPK) and mammalian target of rapamycin (mTOR) have been sug- gested to regulate ULK1 activity, with activation of ULK1 by AMPK through phosphorylation of ULK Ser317 and ULK Ser555 and inhibition of ULK1 by mTOR through phosphorylation of ULK Ser757 (Egan et al. 2011; Kim et al. ª2018 The Authors.Physiological Reportspublished by Wiley Periodicals, Inc. on behalf of The Physiological Society and the American Physiological Society. This is an open access article under the terms of the Creative Commons

by AMPK through phosphorylation of ULK Ser317 and ULK Ser555 and inhibition of ULK1 by mTOR through phosphorylation of ULK Ser757 (Egan et al. 2011; Kim et al. ª2018 The Authors.Physiological Reportspublished by Wiley Periodicals, Inc. on behalf of The Physiological Society and the American Physiological Society. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.2018 | Vol. 6 | Iss. 7 | e13651 Page 1 Physiological Reports ISSN 2051-817X

2011). One mitophagy model suggests a direct interaction between microtubule-associated protein 1A/1B-light chain 3 (LC3) protein on the autophagosome and the mitophagy receptor BCL2 interacting protein (BNIP)3 on the mito- chondria (Kubli and Gustafsson 2012). Another model suggests that the E3 ubiquitin ligase protein Parkin recog- nizes and facilitates ubiquitination of proteins on compro- mised mitochondria (Kirkin et al. 2009). The autophagy adapter protein p62 links the ubiquitinated mitochondria and the autophagosomes together by interaction with LC3II on the autophagosome membrane, promoting elon- gation and enwrapping cytosolic cargos including mito- chondria (Mizushima and Komatsu 2011). However, the effects of exercise and exercise training on these compo- nents in human skeletal muscle are not fully elucidated. The previous observations that AMPK Thr172 and ULK Ser317 phosphorylation both increased, while ULK Ser757 phospho- rylation decreased in human skeletal muscle in response to acute exercise, suggest that exercise enhances autophagy (Schwalm et al. 2015). Similarly, a positive correlation between AMPK Thr172 and ULK Ser555 phosphorylation has been reported, with increased mTOR phosphorylation and unchanged ULK Ser757 phosphorylation in human skeletal muscle in response to 60 min of moderate intensity exer- cise (Møller et al. 2015). Moreover, Beclin1 and BNIP3 mRNA has been shown to increase in human skeletal mus- cle in response to acute ultra-endurance exercise (Jamart et al. 2012). The protein content of p62 has been reported to decrease in both mouse (He et al. 2012; Pagano et al. 2014; Brandt et al. 2017a) and human (Schwalm et al. 2015) skeletal muscle following an acute endurance exercise bout, although others have observed unchanged p62 con- tent after endurance exercise (Jamart et al. 2012; Møller et al. 2015; Fritzen et al. 2016a; Halling et al. 2016). More- over, previous mouse studies have shown an increase in LC3II protein in response to an acute bout of endurance exercise (Grumati et al. 2011; Jamart et al. 2012; Vain- shtein et al. 2015; Halling et al. 2016; Brandt et al. 2017a), whereas the response in human skeletal muscle may be more complex. Hence, a previous study reported a marked increase in LC3b mRNA content in human

an increase in LC3II protein in response to an acute bout of endurance exercise (Grumati et al. 2011; Jamart et al. 2012; Vain- shtein et al. 2015; Halling et al. 2016; Brandt et al. 2017a), whereas the response in human skeletal muscle may be more complex. Hence, a previous study reported a marked increase in LC3b mRNA content in human skeletal muscle in response to acute ultra-endurance running (Jamart et al. 2012), while other studies with shorter exercise duration observed a decrease in LC3II protein content in human skeletal muscle (Møller et al. 2015; Schwalm et al. 2015; Fritzen et al. 2016a). Furthermore, Masschelein et al. (2014) reported no change in LC3II protein content follow- ing acute endurance exercise. Together, this may indicate that the LC3I and LC3II regulation in human skeletal mus- cle depends on the exercise protocol. Mouse skeletal muscle Beclin1 protein has been shown to increase with swimming exercise training, while the BNIP3 and Parkin protein content was unchanged (Ju et al. 2016). On the other hand, exercise training in mice has also been reported to increase both BNIP3 and Parkin protein content in skeletal muscle (Brandt et al. 2017b). Furthermore, in LC3II protein increased, while LC3I and p62 protein content was unchanged in mouse skeletal muscle (Brandt et al. 2017a), with exercise training in one study, whereas LC3I and p62 protein increased, and LC3II did not change in mouse skeletal muscle with exercise training in another study (Brandt et al. 2017b). In addi- tion, exercise training responses in p62 and BNIP3 pro- tein in mouse skeletal muscle have been reported to depend on muscle type, because LC3I, LC3II, and BNIP3 increased and p62 protein content decreased in plantaris mucle, whereas LC3I, LC3II, and p62 all increased, and BNIP3 protein was unchanged in soleus muscle with exer- cise training (Lira et al. 2013). It is possible that the different observations on autop- hagy in response to exercise in human skeletal muscle are due to differences in exercise intensity as has been sug- gested in both humans (Schwalm et al. 2015) and mice (Brandt et al. 2017a). However,

and BNIP3 protein was unchanged in soleus muscle with exer- cise training (Lira et al. 2013). It is possible that the different observations on autop- hagy in response to exercise in human skeletal muscle are due to differences in exercise intensity as has been sug- gested in both humans (Schwalm et al. 2015) and mice (Brandt et al. 2017a). However, the effect of exercise training on the regulation of autophagy in human skeletal muscle as well as the impact of sprint bouts on the regu- lation of autophagy with acute exercise and exercise train- ing using a matched total work approach remains to be elucidated. Therefore, the aim of this study was to investi- gate the effect of exercise and exercise training on the reg- ulation of autophagy in human skeletal muscle and examine the impact of sprint bouts on these responses. Methods Subjects A total of fourteen healthy, recreationally physically active male subjects volunteered to participate in the study. However, samples from only twelve subjects are used in the present analysis (n=6 in each group), because one subject lacked biopsy material to complete all the analysis and another did not complete the intervention. The aver- age age of the subjects was 25 year (ranging from 19 to 33 year) and the average BMI was 23.9 (ranging from 17.5 to 29.3). The subjects were engaged in 1–3 weekly training sessions (team sports, endurance and/or strength training) but were not involved in regular competition. Participants were fully informed of the experimental pro- cedures of the study and written consent was obtained prior to the study. All study procedures were approved by the Ethics Committee of Copenhagen and Frederiksberg communities and adhere to the principles of the Declara- tion of Helsinki Title 45, U.S. Code of Federal Regula- tions, Part 46, Protection of Human Subjects, Revised June 23, 2005, effective June 23, 2005. 2018 | Vol. 6 | Iss. 7 | e13651 Page 2 ª2018 The Authors.Physiological Reportspublished by Wiley Periodicals, Inc. on behalf of The Physiological Society and the American Physiological Society. Exercise-Induced Regulation of Autophagy in Human Muscle N. Brandtet

45, U.S. Code of Federal Regula- tions, Part 46, Protection of Human Subjects, Revised June 23, 2005, effective June 23, 2005. 2018 | Vol. 6 | Iss. 7 | e13651 Page 2 ª2018 The Authors.Physiological Reportspublished by Wiley Periodicals, Inc. on behalf of The Physiological Society and the American Physiological Society. Exercise-Induced Regulation of Autophagy in Human Muscle N. Brandtet al.

Preliminary testing Prior to the study all subjects underwent an incremental test to exhaustion on an electronically braked cycle ergome- ter (Monark 839E, Vansbro, Sweden) to ensure that they had a maximal oxygen uptake (VO 2-max)>45 mL kg 1 min 1 . The subjects cycled 5-min at 125 W and 200 W followed by an 25 W min 1 increase in workload until volitional exhaustion defined as pedaling frequency dropping below 50 rpm. In addition, subjects completed a screening, familiarization, and 45 min time-trial test (TT). Experimental protocol Based on ranked scores from VO 2-max and TT perfor- mance, subjects were ranked 1-14 and were matched in pairs (1st with 2nd, 3rd with 4th etc. until 13th with 14th, and randomly assigned into a continuous moderate intensity training group (MOD;n=6) or a training group with inclusion of sprint intervals (SPRINT;n=6). Subjects completed an 8-week training intervention of 60 min of cycling 39per week. During each training ses- sion, the MOD exercise training group cycled for 60 min at a constant power output 157 20 W eliciting~60% of VO 2-max. The SPRINT training group performed in each exercise training session cycling for 60 min with an average power output of (157 20 W) eliciting~60% of VO 2-max interspersed by six 30-sec sprints (473 79 W) every 10 min followed by cycling at 102 17 W for 3:24 min (Fig. 1). Before and after the 8-week interven- tion period, subjects completed three experimental days separated by at least 48 h. Figure 1.Schematic presentation of the two exercise training protocols consisting of 60 min of cycling (A) continuous cycling with an average power output of 157 20 W eliciting~60% of VO 2-max or (B) continuous cycling for 60 min with an average power output of 157 20 W eliciting~60% of VO 2-max interspersed by six 30-sec sprints (473 79 W) every 10 min followed by cycling at 102 17 W for 3:24 min. On experimental days, muscle biopsies were taken at rest as well as immediately, and 120 min after exercise. ª2018 The Authors.Physiological Reportspublished by Wiley Periodicals, Inc. on behalf of The

output of 157 20 W eliciting~60% of VO 2-max interspersed by six 30-sec sprints (473 79 W) every 10 min followed by cycling at 102 17 W for 3:24 min. On experimental days, muscle biopsies were taken at rest as well as immediately, and 120 min after exercise. ª2018 The Authors.Physiological Reportspublished by Wiley Periodicals, Inc. on behalf of The Physiological Society and the American Physiological Society. 2018 | Vol. 6 | Iss. 7 | e13651 Page 3 N. Brandtet al. Exercise-Induced Regulation of Autophagy in Human Muscle

During the first and last training session muscle biopsies were obtained. Subjects reported to the laboratory in the morning 2 h after consumption of a self-chosen standard- ized breakfast (the same before each trial). During the experimental day, the subjects were allowed to drink water ad libitum but with no food ingestion. A resting biopsy (Pre/Pre Training) was obtained from muscle vastus later- alis followed by completion of the first training session as described for MOD and SPRINT. In addition, a resting muscle biopsy was obtained after 8 weeks of exercise train- ing (Post Training). Muscle biopsies were also obtained immediately (0 h) after exercise and at 2 h of recovery during the first and the last training session. The biopsies obtained after (0 and 2 h) the last training session were obtained for another purpose than the current study and therefore not included in the present analyses. Muscle biopsy samples were immediately frozen in liquid nitrogen and stored at 80°C until further analyses. Blood samples were obtained, but not used in the current study. Muscle analyses A part (~80 mg) of the muscle sample was freeze-dried for at least 48 h and dissected free of visible blood and connective tissue. Dissection was performed under a stereo microscope with an ambient temperature of~18°C and a relative humidity below 30%. After dissection, mus- cle tissue was weighed for lysate preparation. Muscle lysates A part of the muscle sample (~5mg d.w.) was used for lysate generation. The muscle tissue was homogenized in ice-cold buffer (10% glycerol, 20 mmol/L Na-pyropho- sphate, 150 nmol/L NaCL, 50 mmol/L Hepes, 1% NP-40, 20 mmol/Lb-glycerophosphate, 10 mmol/L NaF, 1 mmol/ L EDTA, 1 mmol/L EGTA, 20lg/mL aprotinin, 10lg/mL leupeptin, 2 mmol/L Na 3VO 4, 3 mmol/L benzamidine, pH 7.5) for 2 min at 30 oscilliations per second in a Tis- sueLyser (TissueLyser II, iagen, Valencia, CA, USA). The samples were set to rotate end over end for 1 h at 4°C fol- lowed by centrifugation and the lysates were collected as the supernatant. The protein content in the lysate was determined by the bicinchoninic acid method (Pierce

pH 7.5) for 2 min at 30 oscilliations per second in a Tis- sueLyser (TissueLyser II, iagen, Valencia, CA, USA). The samples were set to rotate end over end for 1 h at 4°C fol- lowed by centrifugation and the lysates were collected as the supernatant. The protein content in the lysate was determined by the bicinchoninic acid method (Pierce Chem, Rockford, IL, USA) and lysates were prepared with sample buffer containing Sodium Dodecyl Sulfate (SDS). SDS-PAGE and western blotting Phosphorylation levels and protein content were measured by SDS-PAGE and western blotting using self-casted gels. PVDF membranes were blocked in 3% fish gel, and protein and phosphorylation sites were determined using primary antibodies against AMPK Thr172 phosphorylation (#2535S, Cell Signaling), AMPKa2 protein (#G3013, Santa Cruz Biotechnology), Beclin1 protein (#3738, Cell Signaling), BNIP3 protein (#12396, Cell Signaling), LC3A/B protein (#4108, Cell Signaling), mTOR Ser2448 phosphorylation (#2971, Cell Signaling), mTOR protein (#2972, Cell Signal- ing), OXPHOS protein (#110413, Abcam), p62 protein (#5114, Cell Signaling), Parkin protein (#4211S, Cell Sig- naling), ULK1 protein (#8054, Cell Signaling), ULK Ser317 phosphorylation (#12753, Cell Signaling) and ULK Ser757 phosphorylation (#6888, Cell Signaling). The membranes were incubated in species-specific horse radish peroxidase- conjugated secondary antibodies (Dako, Glostrup, Den- mark) and protein and phosphorylation levels were visual- ized using LuminataTM Classico Western HRP Substrate (Millipore, Denmark). The OXPHOS protein analyses were performed before heating of the samples to prevent that Complex IV was affected. Band intensity was quantified using ImageQuant Las 4000 (GE Healthcare, Munich, Ger- many) and ImageQuant Imaging software. Protein content and phosphorylation levels were expressed in arbitrary units normalized to control samples loaded on each side of each gel. Statistics Phosphorylation levels and protein content in response to acute exercise and exercise training were evaluated using a two-way analysis of variance (ANOVA) for repeated measures. If a main effect was observed, a Student-New- man-Keul‘s post hoc test was used to locate differences between time points and training protocols. A significance level ofP<0.05 was chosen, and a tendency is reported at 0.05≤P<0.1. Statistical calculations were performed using SigmaPlot Version 12.5. Results Performance Exercise training had no

using a two-way analysis of variance (ANOVA) for repeated measures. If a main effect was observed, a Student-New- man-Keul‘s post hoc test was used to locate differences between time points and training protocols. A significance level ofP<0.05 was chosen, and a tendency is reported at 0.05≤P<0.1. Statistical calculations were performed using SigmaPlot Version 12.5. Results Performance Exercise training had no effect on whole body VO2max in either group (53.8≤5.5 and 54.8≤4.8 mL min 1 kg 1 before and 53.1≤8.3 and 52.3≤7.6 mL min 1 kg 1 after exercise training in MOD and SPRINT, respectively). How- ever, the work performed during the time trial increased similarly in MOD and SPRINT with exercise training (be- fore: 202≤27 and 217≤37 J/sec and after: 220≤30 and 226≤38 J/sec in MOD and SPRINT, respectively). Acute exercise In both MOD and SPRINT, skeletal muscle AMPK Thr172 phosphorylation was~2–3 fold higher (P<0.05) 2018 | Vol. 6 | Iss. 7 | e13651 Page 4 ª2018 The Authors.Physiological Reportspublished by Wiley Periodicals, Inc. on behalf of The Physiological Society and the American Physiological Society. Exercise-Induced Regulation of Autophagy in Human Muscle N. Brandtet al.

The study investigates the effects of exercise on autophagy in human skeletal muscle.