Higher glucose availability augments the metabolic responses of the C2C12 myotubes to exercise-like electrical pulse stimulation

Abstract

The application of exercise-like electrical pulse simulation (EL-EPS) has become a widely used exercise mimetic in vitro. EL-EPS produces similar physiological responses as in vivo exercise, while less is known about the detailed metabolic effects. Routinely, the C2C12 myotubes are cultured in high-glucose medium (4.5 g/L), which may alter EL-EPS responses. In this study, we evaluate the metabolic effects of EL-EPS under the high- and low-glucose (1.0 g/L) conditions to understand how substrate availability affects the myotube response to EL-EPS. The C2C12 myotube, media, and cell-free media metabolites were analyzed using untargeted nuclear magnetic resonance (NMR)-based metabolomics. Furthermore, translational and metabolic changes and possible exerkine effects were analyzed. EL-EPS enhanced substrate utilization as well as production and secretion of lactate, acetate, 3-hydroxybutyrate, and branched-chain fatty acids (BCFAs). The increase in BCFAs correlated with branched-chain amino acids (BCAAs) and BCFAs were strongly decreased when myotubes were cultured without BCAAs suggesting the action of acyl-CoA thioesterases on BCAA catabolites. Notably, not all EL-EPS responses were augmented by high glucose because EL-EPS increased phosphorylated c-Jun N-terminal kinase and interleukin-6 secretion independent of glucose availability. Administration of acetate and EL-EPS conditioned media on HepG2 hepatocytes had no adverse effects on lipolysis or triacylglycerol content. Our results demonstrate that unlike in cell-free media, the C2C12 myotube and media metabolites were affected by EL-EPS, particularly under high-glucose condition suggesting that media composition should be considered in future EL-EPS studies. Furthermore, acetate and BCFAs were identified as putative exerkines warranting more research.

NEW & NOTEWORTHY The present study examined for the first time the metabolome of 1) C2C12 myotubes, 2) their growth media, and 3) cell-free media after exercise-like electrical pulse stimulation under distinct nutritional loads. We report that myotubes grown under high-glucose conditions had greater responsiveness to EL-EPS when compared with lower glucose availability conditions and increased media content of acetate and branched-chain fatty acids suggests they might act as putative exerkines warranting further research.

INTRODUCTION

Adequate physical activity is known to prevent and treat many diseases, such as metabolic, cardiovascular, and musculoskeletal disorders (1). During exercise, muscles secrete molecules that can act as intra- (autocrine and paracrine) or intertissue (endocrine) signaling factors (2). These muscle-derived signaling mediators have been recently shown to promote, for instance, muscle-liver cross talk during and after exercise, which is essential for many physiological processes, such as regulation of energy metabolism during increased fuel demand (3, 4). Overall, recognition of the skeletal muscle as a secretory organ (5, 6) has opened a new research area in exercise physiology.

Pedersen et al. (7) originally named muscle-originated proteins and cytokines as myokines. Afterward, Tarnopolsky and coworkers (8) defined exerkines as myokines and other molecules, such as metabolites, extracellular vesicles, and nucleic acids, secreted from the contracting muscles [i.e., myometabokiome (9)] and other tissues. Due to the large size (30%–40% of the body mass) and great vascularization, contribution of the skeletal muscle to the secreted myokine/exerkine pool is significant (10). A number of in vivo studies have been conducted including analyses of a variety of body fluids (11–13) and muscle tissues as well as examination of arteriovenous difference (14) to examine muscle-derived molecules during rest and exercise. However, these analyses will include molecules secreted from other organs in the body so that metabolic products of skeletal muscle cannot be specified.

To look specifically at skeletal muscle metabolism with exercise, we have used a widely adopted cell culture model to mimic in vivo exercise in vitro. This model involves treating the C2C12 myotubes with exercise-like electrical pulse stimulation [hereafter EL-EPS as recommended (15)], which has been shown to produce similar physiological responses at transcriptional, translational, and metabolic levels as in vivo exercise (for review, see Refs. 15, 16). A major benefit of the in vitro EL-EPS approach is the ability to selectively and exclusively study myotube metabolism and myotube-derived molecules.

Although nutrition is a critical factor that regulates skeletal muscle response to exercise, in vitro studies have largely overlooked the composition of the media (17). Indeed, a recent study showed that the media composition had a major effect on the analyzed metabolite profiles of different cell lines (18). Thus, to raise awareness of this aspect, we examined the effects of two media containing different amounts of glucose and EL-EPS on myotube metabolism. Because the glucose content in routine cell culture medium may differ from normal/healthy physiological range (19), it is important to determine how the metabolic functioning of myotubes after EL-EPS is affected by the glucose availability.

In the present study, we aimed to assess the effects of EL-EPS and nutritional status (glucose availability) on C2C12 myotube metabolism by conducting untargeted nuclear magnetic resonance (NMR)-based metabolomics analysis of both the cell extract and the media. To roughly estimate whether metabolite uptake or release was occurring, we also analyzed cell-free media controls. The latter was analyzed also after EL-EPS to exclude the possible direct effects of EL-EPS on the media. Altogether, our results show that the glucose availability affected a significant number of the observed metabolic changes in response to EL-EPS suggesting that nutrient availability is indeed a critical factor that should be taken into account in the future studies.

MATERIALS AND METHODS

Cell Cultures

Murine C2C12 myoblasts and human HepG2 hepatocytes were purchased from American Type Culture Collection (Manassas, VA). The myoblasts were grown and differentiated as previously described (20). Briefly, the myoblasts were seeded on 6-well plates (NunclonTM Delta; Thermo Fisher Scientific, Waltham, MA) at a density of ∼12,000 cells/cm². The growth medium (GM) contained high-glucose (HG, 4.5 g/L) Dulbecco’s modified Eagle medium (DMEM, No. BE12-614F, Lonza, Basel, Switzerland), 10% (vol/vol) fetal bovine serum (FBS, No. 10270, Gibco, Rockville, MD), 100 U/mL penicillin and 100 µg/mL streptomycin (P/S, No. 15140, Gibco), and 2 mM l-glutamine (No. 25030, Gibco). The differentiation medium (DM) contained HG DMEM supplemented with 5% (vol/vol) FBS, 100 U/mL, and 100 µg/mL P/S and 2 mM l-glutamine. The HG DM was refreshed every two days, except at day 4 postdifferentiation the cells were acclimatized to low-glucose (LG, 1 g/L, No. BE12-707F, Lonza) DM if the following experiments were conducted in LG conditions. According to the medium provider, the only difference between the DMEMs used is the glucose content. The C2C12 experiments were conducted at days 4–6 postdifferentiation in 2 mL of medium. The cells were tested negative for mycoplasma (MycoSPY, M020-025, Biontex Laboratories GmbH, München, Germany). The HepG2 cells were grown in HG DMEM/Glutamax medium (No. 31266, Gibco) supplemented with 10% (vol/vol) FBS and 100 U/mL and 100 µg/mL P/S. The cells were seeded on 10-cm² dishes (NunclonTM Delta; Thermo Fisher Scientific) at a density of ∼9,000 cells/cm². The HepG2 cells experiments were conducted in 5 mL of serum-free (SF) and antibiotic-free medium. All the cell experiments were performed below passage number 9 (C2C12) or 12 (HepG2) in a humidified environment at 37°C and 5% CO₂.

EL-EPS Protocols for C2C12 Myotubes

Comparable low-frequency EL-EPS protocol as used in the present study has previously been reported to induce similar metabolic and translational changes as in vivo exercise (21–23). According to the studies by Nikolić et al. (16) along with visible contractions verified under a microscope (results not shown), the 24-h chronic low-frequency EL-EPS protocol (1 Hz, 2 ms, 12 V) was chosen. On day 6 postdifferentiation, all C2C12 samples were collected immediately after the cessation of the EL-EPS.

EL-EPS for Metabolomics

On day 5 post C2C12 differentiation, the wells were rinsed with phosphate buffered saline (PBS, No. 10010, Gibco) and serum-free (SF) HG or LG DMEM supplemented with 2 mM l-glutamine was added for 1 h (24). The medium was removed, the wells were rinsed with PBS and fresh SF HG or LG DMEM supplemented with 2 mM l-glutamine was added. The chronic low-frequency EL-EPS was applied by placing the C-Dish carbon electrodes attached to C-Pace EM machine (IonOptix Corporation, Milton, MA) to the wells. To roughly elucidate whether the cells possibly take up or release metabolites, we analyzed the metabolome of the cell-free LG and HG media supplemented with 2 mM l-glutamine. As recommended previously (18), the cell-free media were treated identical to the cell-containing samples as they were also incubated for 24 h with and without EL-EPS (i.e., no cells/no power and no cells/power), n = 3 (Supplemental Table S1 and Fig. S2; all Supplemental material is available at https://doi.org/10.6084/m9.figshare.14376413.v1).

Figure 1.

The principal component analysis (PCA), Venn diagram, and heat map visualizations of the identified metabolites with and without exercise-like electrical pulse stimulation (EL-EPS). A: the PCA score plots. Fold changes were logarithmically transformed (log₂) and pareto scaling was used to create the plots. B: overall, 37 and 34 metabolites were identified from the C2C12 cells and from the media, respectively, and of these 24 metabolites were common among groups. C: the identified metabolites were distributed among four biological groups. The heat map categorization (k-means clustering) of the analyzed metabolites in the cells (D) and in the media (E). The dashed lines cluster the metabolites that respond similarly to EL-EPS or to media glucose content. Heat map coloring is based on z-scores. n = 6–8 samples per group. See MANOVA statistics in Supplemental Table S3 for the significant main and interaction effects of EL-EPS or HG. AAs, amino acids; BCFAs, branched-chain fatty acids; LG/HG, low-/high-glucose condition; LG/HG + EPS, EL-EPS in low-/high-glucose condition; SCFAs, short-chain fatty acids.

EL-EPS for Oleate Oxidation

The cells were first acclimatized to dissolved and albumin-complexed 0.1 mM oleic acid (No. O3008, oleic acid-albumin from bovine serum, Sigma-Aldrich, St. Luis, MO) and 1 mM l-carnitine (C0158, Sigma-Aldrich) in either SF LG or HG DMEM supplemented with 2 mM l-glutamine on the day 4 postdifferentiation. The next day, the electrodes were placed directly to the wells and EL-EPS was applied. The measurement of oleate oxidation was carried out for 2 h at 37°C as previously described (25) with slight modifications. Briefly, at differentiation day 6, after 22 h of stimulation, EL-EPS was paused, the media were collected, and centrifuged for 1 min at 1,000 g before storing at −80°C. The cells were rinsed with PBS and fresh SF HG or LG DMEM supplemented with 2 mM l-glutamine, 0.1 mM oleic acid, 1 mM l-carnitine and 1 µCi/mL [9,10-³H(N)] oleic acid (24 Ci/mmol, NET289005MC, PerkinElmer, Boston, MA) was added. The radiolabeled oleic acid was omitted from the negative controls. The EL-EPS was applied for the remaining 2 h.

Nuclear Magnetic Resonance Spectroscopy

The cell lysates and the experiment media (including cell-free controls) were collected and prepared for the ¹H NMR analysis as described previously (26) with slight modifications. Briefly, samples from three wells were pooled to ensure adequate metabolite concentrations per one ¹H NMR measurement. Media from three wells were mixed with cold methanol (600 μL of sample and 1,200 μL of methanol) and cells were scraped into 200 µL of 90% (vol/vol) 9:1 aqueous methanol/chloroform mixture. The resulting supernatants were stored at −80°C before room temperature lyophilization using vacuum concentrator (Speed Vac plus SC110 A Savant Instruments Inc., Farmingdale, NY) equipped with a vacuum pump (Vacuum pump V-700, Büchi, Flawil, Switzerland) and controller (Vacuum Controller V-850, Büchi). The experiments were replicated independently three times, total n = 6–8 per group.

The samples lyophilized at RT were reconstituted as previously described (26) with slight modifications. In brief, Na₂HPO₄-NaH₂PO₄ buffer (150 mM, pH = 7.4) in 99.8% D₂O (Acros Organics, Thermo Fisher Scientific) containing 0.5 mM 3-(trimethylsilyl) propanesulfonic-d6 acid sodium salt (DSS-d6, IS-2 Internal Standard, Chenomx, Edmonton, Canada) was used for reconstitution. The samples were placed in 3-mm round-bottom NMR sample tubes (Norell Inc., Morganton, NC) for analysis. All the NMR spectra were collected using a Bruker AVANCE III HD NMR spectrometer, operating at 800 MHz ¹H frequency (Bruker Corporation, MA) equipped with a cryogenically cooled ¹H, ¹³C, ¹⁵N triple-resonance probehead. The temperature of the samples was set at 25°C during the measurements. For the ¹H one-dimensional (1-D) NOESY experiments, the free induction decay (FID) was sampled with 133,926 points covering the spectral width of 16,741 Hz, using a relaxation delay of 5 s, acquisition time of 4 s, and mixing time of 0.1 s. The signal was accumulated with 128 scans. The obtained data were analyzed using Chenomx 8.5–8.6 software (Chenomx). In addition to ¹H 1-D spectra, heteronuclear ¹H-¹³C single quantum coherence spectroscopy (HSQC) and ¹H-¹³C HSQC-total correlation spectroscopy (HSQC-TOCSY), as well as homonuclear ¹H-¹H TOCSY and ¹H-¹H double quantum filtered correlation spectroscopy (DQF-COSY) two-dimensional (2-D) spectra were used to confirm the identification of the profiled metabolites. The TopSpin 4.0.9 software (Bruker Corporation) was used for processing and analysis of the 2-D spectra. The spike in-analyses of isobutyric acid (No. I1754, Sigma-Aldrich), isovaleric acid (No. 129542, Sigma-Aldrich) were included.

Oleate Oxidation

After the EL-EPS, the media were run through ion-exchange columns containing Dowex-OH-resin (pH 7, 1X8-200, Cat no. 217425, Sigma Aldrich) (25). Deionized H₂O was used to elute the ³H₂O, which originates from intracellular [9,10-³H(N)] oleic acid β-oxidation that was further secreted to the media. The radioactivity was analyzed as disintegration per minute (DPM) in Optiphase HiSafe 3 scintillation cocktail (Cat. No. 1200.437, PerkinElmer) with Tri-Carb 2910 TR Liquid Scintillation Analyzer (PerkinElmer). The results were calculated using PerkinElmer equations (https://www.perkinelmer.com/fi/lab-products-and-services/application-support-knowledgebase/radiometric/radiochemical-calculations.html). The cells were washed twice with PBS and harvested for total protein content analysis as previously described (20) except for centrifugation at 13,000 g for 10 min at +4°C. The oleate oxidation results were normalized against total protein content and the experiments were replicated independently three times, total n = 8–10 per group.

HepG2 Hepatocyte Experiments

Normal and steatotic HepG2 hepatocytes were used in the experiments. Based on our dose-response experiment steatosis, i.e., fat accumulation, was induced by 24-h administration of 500 µM oleic acid (No. 03008, Sigma-Aldrich) in serum- and antibiotic-free conditions when compared with the nonexposed hepatocytes (Supplemental Fig. S1). The acetate (sodium acetate, CAS No. 127-09-3, Merck, Darmstadt, Germany) dose-response experiment in steatotic hepatocytes suggested that a greater dose (3 mM) that was observed in ¹H NMR analysis (1.5 mM) had no additional effect on intracellular triacylglycerol content over the lower dose (Supplemental Fig. S1).

The normal and steatotic hepatocytes were administered with EL-EPS-stimulated or unstimulated C2C12 conditioned medium (CM) or alternatively with or without 1.5 mM acetate. The C2C12 cells were treated as described for the HG ¹H NMR analysis. After EL-EPS, the media of the stimulated and unstimulated cells were collected and centrifuged for 5 min at 217 g RT to remove cell debris before administration on hepatocytes. In another set of experiments, 1.5 mM acetate or equivalent volume of PBS was administered on both normal and steatotic hepatocytes in serum- and antibiotic-free DMEM/Glutamax. After the 24-h incubation, triacylglycerol extraction from the hepatocytes was conducted. Briefly, the media were collected, centrifuged for 5 min at 217 g, RT and stored at −80°C until use. The HepG2 cells were washed and scraped into PBS, while subsamples for the measurement of total protein content were homogenized into previously described buffer (20). Next, 2:1 methanol-chloroform mixture was added to PBS-cell suspension followed by 5 min centrifugation at 724 g, RT. The supernatant was transferred into a new tube and chloroform, 50 mM citric acid, and H₂O were added. Methanol and chloroform phases were separated by centrifugation for 10 min at 724 g, RT. Chloroform phase was collected and evaporated at +70°C using SpeedVac Concentrator (Thermo Fisher Scientific). The resulting lipid pellet was dissolved into ethanol before measurement. The content of intracellular triacylglycerol as well as glycerol and cytokines in the media were measured as described in the next two sections.

Measurement of Total Protein Content, Enzyme Activities, Triacylglycerol, and Media Glycerol Contents

Total protein content (Bicinchoninic Acid Protein Assay Kit, Pierce Biotechnology, Rockford, IL), triacylglycerol (No. 981786, Thermo Fisher Scientific), and glycerol (No. 984316, Thermo Fisher Scientific) concentrations as well as lactate dehydrogenase (LDH) (No. 981906, Thermo Fisher Scientific) and citrate synthase (CS) (No. CS0720, Sigma-Aldrich) enzyme activities were measured with an automated Konelab or Indiko plus analyzer (Thermo Fisher Scientific). All assays were conducted according to manufacturer’s protocols and enzyme activities in the cells were normalized against total protein content.

4-Plex Cytokine ELISA Analyses

The C2C12 and HepG2 media were centrifuged for 1 min at 1,000 g or 5 min at 217 g, respectively, at +4°C and resulting supernatants were stored at −80°C until use. Next, 25 µL of the samples were directed to mouse [Q-Plex Mouse 4-plex Cytokine Panel (No. 115549MS, Quansys Biosciences, North West, UT)] or human 4-plex Cytokine Panel (Q-Plex Human Cytokine High Sensitivity, No. 112533HU, Quansys Biosciences) assay that were conducted according to the manufacturer’s protocols. In the murine assay, the limit of detection for interleukin-1β (IL-1β) was 12.41 pg/mL, for IL-6 2.90 pg/mL, for tumor necrosis factor α (TNF-α) 3.40 pg/mL, and for interferon γ (IFN-γ) 5.40 pg/mL. In the human assay, the limit of detection for IL-4 was 0.02 pg/mL, for IL-6 0.30 pg/mL, for IL-10 2.39 pg/mL, and for IFN-γ 0.09 pg/mL.

Protein Extraction and Western Blot

The cells were harvested for Western blot and enzyme activity analysis as previously described (20) except for centrifugation at 13,000 g for 10 min at +4°C. The Western blot was conducted as previously described (20). Briefly, 10 µg of total protein per samples were loaded on 4%–20% Criterion TGX Stain-Free protein gels (No. 5678094, Bio-Rad Laboratories, Hercules, CA) and samples were separated by SDS-PAGE. To visualize proteins using stain-free technology, the gels were activated and the proteins were transferred to the PVDF membranes followed by blocking and overnight probing with primary antibodies at +4°C (27). Enhanced chemiluminescence (SuperSignal west femto maximum sensitivity substrate; Pierce Biotechnology, Rockford, IL) and ChemiDoc MP device (Bio-Rad Laboratories) were together used for protein visualization. Stain free (whole lane) was used as a loading control and for the normalization of the results. Primary antibodies used in the present study were purchased from Cell Signaling Technology: p38^{Thr180/Tyr182} (No. 4511), p38 (No. 9212), ERK1/2^{Thr202/Tyr204} (No. 9101), ERK1/2 (No. 9102), SAPK/JNK1/2^{Thr183/Tyr185} (No. 4668), and SAPK/JNK1/2 (No. 9252). The horseradish peroxidase-conjugated secondary IgG antibody was purchased from Jackson ImmunoResearch Laboratories, PA.

Statistical Analyses

The two-way multivariate analysis of variance (two-way MANOVA) was used to analyze main and interaction effects, whereas the group comparisons were conducted by using multivariate Tukey’s test unless stated otherwise (IBM SPSS Statistics, version 26 for Windows, SPSS Chicago, IL). The Spearman’s correlation coefficient was used to analyze correlations (SPSS). The VIsualization and Integration of Metabolomics Experiments (VIIME) software [https://viime.org (28)] was used to generate the heat maps and principal components analyses. The results are presented as means ± SE. The level of significance was set at P < 0.05.

RESULTS

EL-EPS Yielded Different Metabolic Responses under LG and HG Conditions

We studied the effects of chronic low-frequency EL-EPS and medium glucose content on the metabolism of C2C12 myotubes using untargeted ¹H NMR-based metabolomics analysis of the conditioned media and cell extracts. The cell-free media controls were incubated for 24 h with and without EL-EPS. This allowed us to show that EL-EPS does not induce changes in the metabolite profiles in the cell-free media (Supplemental Fig. S2 and Table S1). The principal component analysis (PCA) of the metabolite profiles demonstrated that the four study groups were clearly separated, especially in the media (Fig. 1A). Interestingly, in the PCA of the media, the first principal component separates the groups based on glucose levels and the second principal component separates them based on the application of EL-EPS (Fig. 1A).

The NMR-based metabolomics analysis resulted in identification of 47 individual metabolites. More specifically, we quantified 39 metabolites from the cells and 37 metabolites from the media and the reporting threshold (i.e., the metabolite was detected in over 50% of cases) was met by 37 and 34 metabolites, respectively (Supplemental Table S2). Among the cells and the media, 24 metabolites were shared (Fig. 1B). Overall, the heat map clustering of the metabolites quantified from the cells and media demonstrated that the stimulation-induced differences in the metabolites between LG and HG conditions were greater in the latter, especially in the cells (Fig. 1, D and E). The hierarchical cluster analysis of heat maps from the cell extracts suggests that the metabolites clustered into three categories including those responsive to EL-EPS and to distinct media glucose contents, whereas in the media more categories were observed (Fig. 1, D and E).

Similar to a previous study (12), most of the identified metabolites were distributed among four biological groups. These were 1) metabolism of energy related metabolites (creatine, carbohydrates, and TCA cycle intermediates; 14 metabolites), 2) short- and branched-chain fatty acids (SCFAs and BCFAs, respectively) and ketone bodies (six metabolites), 3) amino acids and related metabolites (24 metabolites) as well as 4) vitamins and others (three metabolites) (Fig. 1C, for individual metabolites, see Supplemental Table S2). In the cells, 18 metabolites were altered due to either EL-EPS or medium glucose content (i.e., EPS and HG main effects, respectively), while eight metabolites demonstrated an interaction effect (EPS × HG) (Supplemental Table S3). In the media, EPS had a main effect on 17 and HG on 13 metabolites, while the interaction effect was detected in seven metabolites (Supplemental Table S3).

Glycolytic ATP Production and Acetate Responded Strongly to EL-EPS

During the EL-EPS, the glucose content in the media decreased when compared with cell-free media indicating increased consumption to support the contraction-induced increase in the energy demand in the cells (Fig. 2A). Simultaneously, lactate content increased both in the cells and in the media, whereas the level of phosphocreatine decreased and dephosphorylated creatine increased suggesting that both glycolytic and phosphocreatine energy sources were utilized (Fig. 2, B–D). In agreement with the increased lactate production and secretion in our experiments, the lactate dehydrogenase (LDH) activity was increased in the cells after EL-EPS, especially in the HG condition (Fig. 2E). Finally, we observed an increase in the cell and media content of acetate, a short-chain fatty acid (SCFA) that can act as a potential fuel source during exercise (29). It appears that in the resting C2C12 cells, the net uptake of acetate was enhanced based on substantially lower acetate content in the cell media than in the cell-free media (LG or HG vs. cell-free LG or HG, Student’s t test, P < 0.001, Fig. 2F). In contrast, during EL-EPS acetate secretion exceeded its uptake partly due to the increased cellular production, at least in HG condition (HG + EPS vs. cell-free HG + EPS, Student’s t test, P < 0.05, Fig. 2F). The effect of the increased glycolysis in response to EL-EPS was accompanied by unaltered content of citrate, the first TCA cycle intermediate, and unaltered citrate synthase (CS) enzyme activity, whereas the levels of the intermediates observed later in the cycle including succinate, fumarate, and malate were increased in the cells in HG condition (Fig. 3, A–E).

Figure 2.

Exercise-like electrical pulse stimulation (EL-EPS) increased energy utilization via glycolytic pathways and readily available energy stores. Glucose (A), lactate (B), dephosphorylated creatine (C), and phosphocreatine contents (D) in the C2C12 cells and/or in the media. E: lactate dehydrogenase (LDH) enzyme activity in the cells. F: acetate content. *P < 0.05; ***P < 0.001, respectively. n = 6–8 samples per group. The two-way MANOVA was used to analyze the effects of the applied EL-EPS and media glucose content (EPS and HG effects, respectively) and their interaction effect, whereas group comparisons were analyzed with multivariate Tukey’s test. In A and F, the dashed lines represent the levels of the cell-free low-glucose (LG) and high-glucose (HG) media controls (i.e., mean of the stimulated and nonstimulated media). Lack of the dashed lines refers to the undetected metabolite content from the cell-free media controls. LG/HG + EPS, EL-EPS in low-/high-glucose condition.

Figure 3.

Increase in the content of tricarboxylic acid (TCA) cycle intermediates in the C2C12 cells after exercise-like electrical pulse stimulation (EL-EPS) was dependent on medium-glucose availability. The contents of citrate (A), succinate (B), fumarate (C), and malate (D). E: citrate synthase (CS) enzyme activity in the cells. *P < 0.05, ***P < 0.001, respectively. n = 6–8 samples per group. The two-way MANOVA was used to analyze the effects of the applied EL-EPS and media glucose content [EPS and high-glucose (HG) effects, respectively] and their interaction effect, whereas group comparisons were analyzed with multivariate Tukey’s test. LG, low-glucose condition; LG/HG + EPS, EL-EPS in low-/high-glucose condition.

Increased Intracellular Amino Acid Levels after EL-EPS

Overall, amino acids were more affected by the EL-EPS than by the glucose availability and only minor effects were observed between HG and LG conditions. Figure 4A shows a forest plot of the amino acids with the log₂ fold changes of the metabolites in response to EL-EPS shown along the x-axis (for individual amino acid box plots, see Supplemental Fig. S3). The plot shows the levels of each of the amino acids under both LG and HG conditions. A set of seven amino acids were increased and eight remained unchanged in response to EL-EPS in the cells, whereas in the media five amino acids increased, two decreased, and ten remained unaltered (Fig. 4B). In contrast, a shared increasing HG effect was observed in three amino acids in the cells and media, whereas a decreasing HG effect was observed in one and two amino acids, respectively (Fig. 4C). The contents of cysteine (Student’s t-test, media vs. cell-free media, P < 0.01), glycine (P < 0.001), histidine (P < 0.05), and lysine (P < 0.001) were lower in the cell-free than in the cell-containing media suggesting release of these amino acids from the C2C12 cells and for cystine and lysine release appears to be further increased during EL-EPS (Supplemental Fig. S3). In contrast, serine (P < 0.001) and glutamine (P < 0.001) contents were greater in the cell-free media controls suggesting active uptake of these amino acids by the C2C12 cells and this appears to be further increased during EL-EPS (Supplemental Fig. S3).

Figure 4.

The effects of exercise-like electrical pulse stimulation (EL-EPS) on the analyzed amino acids were greater than the effect of the glucose availability. Forest plots of the amino acids in the C2C12 cells (A) and in the medium (B) analyzed as logarithmically transformed fold changes [Log₂ (FC)], n = 6–8 samples per group. The error bars demonstrate 95% confidence intervals of the analyzed groups [low-glucose (LG) and EL-EPS in low-glucose condition (LG + EPS) or high-glucose (HG) and EL-EPS in high-glucose condition (HG + EPS)]. Next to the plots are shown the interaction effect (EPS × HG) and the main effects of the applied EL-EPS (EPS) and media glucose content (HG) analyzed by the two-way MANOVA. If the interaction effect is significant, the main effects are shown in the brackets. Multivariate Tukey’s test was used to analyze the group comparisons and significant results are depicted as larger dots. Red = low glucose samples, blue = high glucose samples. Gray arrows depict the direction (increase or decrease) of the main effect. Pie charts of the amino acids demonstrating EPS (B) and HG main effects (C) in the cells and in the medium. Of note: the 95% confidence intervals were calculated based on Student’s t distribution, whereas the group comparisons were conducted using more stringent Tukey’s test.

Increased Intra- and Extracellular Contents of Branched-Chain Fatty Acids after EL-EPS

A set of branched-chain fatty acids (BCFAs) demonstrated significant increases induced by EL-EPS. The levels of 2-methylbutyrate, isobutyrate, and isovalerate were increased in the media after EL-EPS independent of the glucose availability showing their release/secretion from the cells (BCFAs were not detected from the cell-free media) (Fig. 5, A–C). Of these BCFAs, isobutyrate and isovalerate were also increased in the cells after EL-EPS, but this was explained by the increase in HG condition (Fig. 5, A–C). Indeed, the responses of BCFAs to EL-EPS were overall greater in HG condition. The observation of the BCFAs was unanticipated and the source of these metabolites was not entirely clear. Based simply upon chemical structure, we suspected that these metabolites could be the result of branched-chain amino acid (BCAA) catabolism. To test this, we evaluated the correlations between the BCFAs and the BCAAs. As shown in Supplemental Table S4, we found no significant correlations in the media, but very strong correlations were found between the BCAA and BCFAs in the cells. Figure 5D postulates the pathway through which the BCAAs are transformed.

Figure 5.

The production and secretion of branched chain fatty acids (BCFAs) and ketone bodies were increased after exercise-like electrical pulse stimulation (EL-EPS) independent of the glucose availability. The contents of 2-methylbutyrate (A), isobutyrate (B), and isovalerate (C). D: schematic presentation of the branched-chain amino acids and their breakdown metabolites, BCFAs. E: the content of ketone body 3-hydroxybutyrate in the media. *P < 0.05, **P < 0.01, and ***P < 0.001. n = 6–8 samples per group. The two-way MANOVA was used to analyze the effects of the applied EL-EPS and media glucose content (EPS and high-glucose (HG) effects, respectively) and their interaction effect, whereas group comparisons were analyzed with multivariate Tukey’s test E: dashed lines represent the level of the cell-free low-glucose (LG) and HG media controls (i.e., mean of the no cells/no power and co cells/power content). Lack of the dashed lines refers to the undetected metabolite content from the cell-free media controls. ACOT9, acyl-CoA thioesterase 9.

To unequivocally confirm the identity of the BCFAs, we conducted spike-in ¹H NMR experiments, where authentic standards of these compounds were added to the samples to show that the spectral patterns were clearly matching. Finally, we also confirmed by culturing the C2C12 myotubes in BCAA-free media that indeed, BCFAs appear to originate from the BCAA breakdown based on their greater abundance in standard BCAA containing media (pilot results, Supplemental Fig. S3).

Increased Ketone Body Levels in the Media after EL-EPS

The ketone body 3-hydroxybutyrate was identified in the cell-free and the C2C12 media (Fig. 5E). It should be noted that the signals for 3-hydroxybutyrate in the cell-free and unstimulated media samples were near the limit of detection and thus the quantitation is only approximate. Application of EL-EPS led to a significant increase in the signals for 3-hydroxybutyrate in the C2C12 media enabling confident identification and quantitation. The increase in 3-hydroxybutyrate content was greater under HG condition demonstrating that ketone body production in these cells was affected by the glucose availability.

Glucose Availability Resulted in Variable Changes in Exercise and Stress Associated Markers

As the glucose content together with the EL-EPS influenced the metabolite levels, we investigated next whether this was also translated to the phosphorylation levels of the mitogen-activated protein kinases (MAPKs) that are common markers of skeletal muscle after energetic stress and exercise (30). We observed that EL-EPS increased the phosphorylation of stress-activated protein kinase/c-Jun N-terminal kinase (SAPK/JNK)^{Thr183/Tyr185} independent of the glucose availability (Fig. 6A), whereas the phosphorylation of p38^{Thr180/Tyr182} and extracellular regulated kinase (ERK1/2)^{Thr202/Tyr204} remained unaltered (Supplemental Fig. S4). Because SAPK/JNK has been shown to regulate IL-6 signaling in the C2C12 myotubes after EL-EPS (31), we examined the common exercise-responsive cytokines from the media and found that only IL-6 was detectable after EL-EPS. The IL-6 concentration was increased independent of the glucose availability (Fig. 6B). Furthermore, Hojman et al. (32) showed that IL-6 release from the muscle cells is at least in part dependent on the lactate production, and similarly our correlation analysis demonstrated a positive association between the media IL-6 content and the content of lactate in the cells (r = 0.551, P < 0.01) and in the media (r = 0.660, P < 0.001). For the full list of the IL-6 and metabolite correlations, see Supplemental Table S4.

Figure 6.

Exercise-like electrical pulse stimulation (EL-EPS) promoted protein phosphorylation and cell metabolism especially under high-glucose (HG) condition. A: phosphorylated stress-activated protein kinase/c-Jun N-terminal kinase (SAPK/JNK1/2)^{Thr183/Tyr185}, total SAPK/JNK1/2, and representative blots. −, no stimulation; +, stimulation. In the figure, the values are presented as normalized to low-glucose (LG) = 1 or HG = 1. Interleukin-6 (IL-6) concentration in the media (B), oleate oxidation rate (C), intracellular triacylglycerol (TG) content (D), and media glycerol content (E). *P < 0.05, **P < 0.01, and ***P < 0.001. In A and D–E, n = 6, in B, n = 12 [pool of the ¹H NMR and oleate oxidation (22-h timepoint) experiments] and in C, n = 8–10 samples per group. The two-way MANOVA was used to analyze the effects of the applied EL-EPS and media glucose content (EPS and HG effects, respectively) and their interaction effect, whereas group comparisons were analyzed with multivariate Tukey’s test.

The changes in the metabolites in response to EL-EPS suggest that the C2C12 cell line is very glycolytic in nature even under low-frequency stimulation and therefore in addition to the unaltered activity of the citrate synthase as a TCA cycle marker (Fig. 3E), we expected that the fatty acid oxidation might not increase. To test this hypothesis, oleic acid and l-carnitine acclimatized C2C12 myotubes were applied with 24-h EL-EPS during which radiolabeled [9,10-³H(N)] oleic acid was added together with unlabeled oleic acid and l-carnitine in fresh media. The rate of oleate oxidation was analyzed as the amount of ³H₂O produced and secreted by the myotubes to the culture media. Indeed, the analysis of the metabolic effects of EL-EPS demonstrated that oleate oxidation even decreased in the cells under HG condition (Fig. 6C). This occurred without EL-EPS-induced changes in triacylglycerol content, although media glycerol content as a marker of lipolysis was increased in HG condition after EL-EPS (Fig. 6, D and E).

EL-EPS Has No Effect on Cell Viability but Myotubes Grown in HG May Be More Viable

To understand whether the myotubes were more viable and thus perhaps more metabolically active in the HG condition, which could partly explain some of the observed results, LDH enzyme activity was measured from the media as a marker of cell rupture (16) (Supplemental Fig. S5). Overall, the cell viability remained unaffected by the EL-EPS protocol. However, LDH activity tended to be lower in HG condition in NMR experiments thus suggesting possibly better viability than in LG condition (Supplemental Fig. S5).

Potential Myotube-Derived Exerkines Had Little Effect on Normal and Steatotic Hepatocytes

An initial goal of this study was to search for potential, myotube-derived metabolites that could act as exerkines. To test whether myotube-derived exerkines can alter hepatic steatosis, we cultured HepG2 hepatocytes and induced the accumulation of triacylglycerol by supplementation of the media with oleic acid (Supplemental Fig. S1). Steatosis was not accompanied with inflammation because inflammatory markers (IL-4, IL-6, IL-10, and INF-γ) in the media remained below the detection limit (data not shown). As acetate was the metabolite secreted with the largest fold change in response to EL-EPS, we administrated normal and steatotic hepatocytes with 1.5 mM acetate or with EL-EPS CM. The acetate concentration was chosen based on the ¹H NMR analysis results. After the 24-h incubation, we observed that neither of the approaches had adverse effects on the triacylglycerol content of the hepatocytes or the glycerol content in the media (Fig. 7, A–D). That said, the level of glycerol in the media as a marker of lipolysis approached an increasing trend after EL-EPS CM administration (EPS main effect, P = 0.098, Fig. 7D).

Figure 7.

The 24-h administration of acetate and conditioned medium from the stimulated C2C12 cells had only minor effects on normal and steatotic HepG2 hepatocytes. Cell triacylglycerol (A) and media glycerol (B) contents after administration of 1.5 mM acetate (diluted in PBS) or PBS control. Cell triacylglycerol (C) and media glycerol (D) contents after administration of the media from the stimulated or unstimulated C2C12 cells grown under high-glucose conditions [electrical pulse stimulation (EPS) and CTRL medium, respectively]. n = 3 samples per group. The two-way MANOVA was used to analyze the effects of the applied exercise-like (EL)-EPS and steatosis (EPS and health effects, respectively) and their interaction effect, whereas group comparisons were analyzed with multivariate Tukey’s test.

DISCUSSION

Skeletal muscle metabolism is known to increase dramatically from rest to exercise and the ability of the cells to adapt to increased energy demand is vital (33). In agreement with these known in vivo physiological facts and similar to previous in vivo studies (11, 12, 34), we observed a number of perturbations to energy metabolism that were affected by EL-EPS. Our studies also observed a significant impact of nutrient availability on the EL-EPS-induced metabolic changes. Most, but not all of the EL-EPS responses were of larger magnitude in high-glucose conditions, which may be due to the fact that in low-glucose condition the 24-h EL-EPS almost completely depleted media and cells from glucose. The decreased glucose content in the media after EL-EPS is in line with previous studies demonstrating that EL-EPS promotes glucose uptake into the myotubes (16) and the well-known fact that exercise in vivo increases glucose uptake into the skeletal muscle (35). Similarly, in agreement with in vitro (16) and in vivo (36) findings, we observed an increased production and secretion of lactate and decreased intracellular content of phosphocreatine demonstrating that the applied EL-EPS induced anaerobic and especially glycolytic ATP production and energy demands in the C2C12 myotubes. Indeed, the C2C12 cell line has been considered to be glycolytic in nature (37) and rely on anaerobic glycolysis at rest (38), which may explain the increased lactate production and decreased fat oxidation during the applied low-frequency EL-EPS. As lactate plays an important role in intercellular signaling of nearby and/or distant cells (39), its role as an exerkine has probably been underappreciated and should be further studied.

Our studies revealed alterations in a set of short- and branched-chain fatty acids (S/BCFAs) and ketone bodies with EL-EPS. SCFAs such as acetate, propionate, and butyrate are commonly observed with in vivo studies and are common products of gut microbial metabolism (40), while also other tissues can produce SCFAs. For example, Van Hall et al. (41) showed that exercise increased leg acetate release by ninefold when compared with rest. In the present study, we showed that intra- and extracellular contents of acetate were increased after EL-EPS, while the former was more pronounced in high-glucose condition. During increased energy demand or restricted TCA cycle function, all of the pyruvate-derived acetyl-CoA might have not successfully entered the TCA cycle and thus the excess can be hydrolyzed to acetate and released into the circulation (42, 43). Acetate can also be produced from pyruvate via enzymatic and nonenzymatic reactions including pyruvate dehydrogenase (PDH) and reactive oxygen species (ROS) (44), respectively, and PDH activity (45) and ROS levels (16) are increased by exercise and/or myotube contractions. In addition, hyperactive glucose metabolism and nutritional excess have been related to incomplete metabolism and excretion of metabolites, which lead to promoted conversion of pyruvate to acetate (44). This might have been the case also in our study since in high-glucose condition, the accumulation of succinate, fumarate, and malate suggests that substrate availability and entry to the TCA cycle might have exceeded the capacity of the oxidative phosphorylation machinery thus resulting in incomplete substrate oxidation. It is possible that the reduced need for full TCA cycle activity is due to enhanced glycolysis, which may provide enough ATP for the working myotubes based on the strongly increased lactate levels.

As acetate has been shown to positively modify liver lipid metabolism (46), the effects of myotube-derived acetate and the whole EPS secretome (EL-EPS CM) on hepatocytes were examined. By applying acetate and EL-EPS CM to normal and steatotic hepatocytes, we found that the intracellular triacylglycerol content in the hepatocytes remained unaltered in the studied conditions. However, the molecules originated from the contracted muscle cells (EL-EPS CM) had a tendency for increased lipolysis in the hepatocytes, inferred from the increased media content of glycerol. In long-term, this could lead to reduced intracellular triacylglycerol content, but further studies are needed. Also more studies investigating dose-response effects of acetate are also warranted. This is because we found high levels of acetate from the cell-free media controls similar to a previous study (47), meaning that hepatocyte culture already had high levels of acetate before further adding it into the media. Future physiology studies should also investigate whether the high levels of acetate use and release from muscle cells have physiological effects in vivo.

Besides acetate, the media content of 3-hydroxybutyrate, a common ketone body, increased after EL-EPS and the response was greater in high-glucose condition. Previously, an increased content of circulating 3-hydroxybutyrate after exercise has been considered to act as a biomarker of metabolic shift from the utilization of carbohydrates toward fats (34) and SCFAs and ketone bodies have been reported to positively modify lipid, carbohydrate, and protein metabolism in the muscle and in other tissues (48). Although 3-hydroxybutyrate has been considered to be produced mainly by the liver, the growing body of evidence demonstrates that during exercise skeletal muscle could secrete certain ketone bodies and thus contribute to the circulating ketone body pool (13). The 3-hydroxybutyrate has been detected in the muscle interstitial fluid after exercise (13), and the enzyme regulating 3-hydroxybutyrate synthesis, HMG-CoA synthase (HMGCS2), has been shown to be elevated in skeletal muscle after exercise (49). In addition, 3-hydroxybutyrate dehydrogenase (BDH1), which converts acetoacetate into 3-hydroxybutyrate has been shown to be increased in the skeletal muscle by exercise and decreased by inactivity (49) (https://metamex.com). Thus, 3-hydroxybutyrate could be an exerkine, however, further studies are needed to verify its functions on nearby cells and the whole body metabolism and cross talk after exercise.

In addition to the more routinely observed SCFAs, we observed a set of unique changes in several BCFAs including 2-methylbutyrate, isobutyrate, and isovalerate that all increased after EL-EPS. Correlations between BCFAs and BCAAs along with studies using BCAA-depleted media strongly indicate that the BCFAs are indeed derived from BCAA catabolism. Previous in vivo study suggested that during exercise BCFA precursors derived from the BCAA catabolism were increased in the skeletal muscle (50) and identified from the circulation (51). Concordant with this finding, we demonstrated that BCFAs can be produced and released by muscle cells in response to EL-EPS. Furthermore, we simultaneously observed an increase in the BCAAs in the cells after EL-EPS, identical as reported in glycolytic human type II-muscle fibers after exercise (52). Previous studies have shown that increased levels of circulating BCAAs and decreased BCAA degradation has been associated with poor metabolic health (53). In this study, we reported 1) unaltered media BCAA content, 2) increased content of BCAA breakdown products, and 3) increased intracellular content of many amino acids (including BCAAs) after EL-EPS. Together these results suggest that the EL-EPS perhaps enhanced protein breakdown and amino acid recycling similarly as in vivo exercise (54). In summary, high correlations of the BCAAs and their breakdown products as well as very low levels of BCFAs in the experiment with BCAA-depleted media support the evidence that in the C2C12 cells the origin of BCFAs seems to be BCAA catabolism.

The enzymes needed for the conversion of the acyl-CoA derivatives originated from the BCAA catabolism [2-methylbutyryl-CoA, isobutyryl-CoA, and isovaleryl-CoA (55)] into 2-methylbutyrate, isobutyrate, and isovalerate may include acyl-CoA thioesterases (ACOTs), of which skeletal muscle expresses different isoforms (56) and the C2C12 cell line at least Acot3 and Acot9 (57). The ACOT9 isoform has been shown to have a unique substrate specificity with the ability to hydrolyze short-chain acyl-CoA esters, including isobutyryl-CoA and isovaleryl-CoA (56, 58). This further suggests that the mitochondrial link between branched-chain fatty acid and amino acid metabolism could be ACOT9 (58).

The effect of EL-EPS on the amino acids was consistently greater than the high-glucose effect both in the cells and in the media. The in vivo changes in the circulating amino acids have been reported to be controversial and dependent on their specific glucogenic versus ketogenic, nonessential versus essential or other properties (11, 12) as well as exercise intensity (59). To summarize, although our results are mainly in agreement with recent in vivo systematic reviews on exercise metabolomics (11, 12), more research is needed to better understand the stimulation-induced changes in myotube and media metabolites. Indeed, the number of studies analyzing the metabolome after EL-EPS is small (60) and, to best of our knowledge, this study is the first to study metabolites after EL-EPS under distinct nutritional loads.

We observed a decline in oleate oxidation during EL-EPS under high-glucose condition, whereas under low-glucose condition the decrease approached a trend. Decreased oleate oxidation may be related to the aforementioned 1) incomplete oxidation of the energetic intermediates during hyperactive metabolism and nutrient overloading (44) and 2) glycolytic nature of the C2C12 cell line (37). Addition of the fresh media containing the radiolabeled and unlabeled oleic acid in right proportion is an essential step for accurate oxidation measurement when using our protocol. Based on our results, the 24-h EL-EPS almost completely depleted the glucose from the media in LG condition. In oleate oxidation experiments, we had to replace the EL-EPS media with the fresh ³H oleic acid-containing media after 22 h of stimulation and thus, we simultaneously provided the cells with fresh glucose for the remaining 2 h of the EL-EPS. This may have temporarily stimulated the cells to rely heavily on glucose, perhaps, at least in LG condition. Furthermore, increased glycolysis causes accumulation of acetyl-CoA that could inhibit β-oxidation via downregulation of β-ketoacyl-CoA thiolase (61). That said, because acetyl-CoA cannot be transported across membrane, the excess might also be transformed to acetate via the action of ACOT enzymes (58) or via nonenzymatic processes (44) and then secreted from the cells to balance the intracellular state of TCA cycle substrates (42) as we observed in the present study. Besides decreased oleate oxidation, we observed no changes in intracellular triacylglycerol content similar to the work by Laurens et al. (62), although our observation of the increased glycerol release may be associated with enhanced lipolysis under high-glucose condition after EL-EPS. In addition, contraction-induced changes in triacylglycerol content may be pretreatment specific because acclimatization of the C2C12 myotubes to fatty acids has been shown to cause intramyocellular triacylglycerol accumulation, which was prevented by short-duration low-frequency EL-EPS (63). Moreover, previous studies have reported controversial results on fatty acid oxidation after EL-EPS and the main factors affecting the rate of β-oxidation seem to be related to the stimulation protocol, duration of the measurement, fatty acid analyzed (e.g., oleate or palmitate), analysis protocol/method, and the cell line used (37, 38, 63–66).

In addition to changes in the metabolome, the applied EL-EPS altered the phosphorylation and secretion of stress-inducible markers, such as MAPKs and cytokines, commonly analyzed after in vivo exercise and in vitro EL-EPS (16). The phosphorylation of SAPK/JNK increased after EL-EPS independent of glucose availability, which is supported by in vivo studies demonstrating that glycolytic exercise increases SAPK/JNK phosphorylation (67, 68). Of the exercise-responsive cytokines, we observed an increase in the secretion of IL-6 after EL-EPS, also independent of glucose availability. Interestingly, muscle IL-6 secretion may occur in part through proteasome-dependent release initiated by lactate production (32). Indeed, the IL-6 in the media correlated positively with the intra- and extracellular lactate content in the present study (Supplemental Table S4). In addition, SAPK/JNK has been shown to regulate IL-6 metabolism (31) and indeed we observed that IL-6 and phosphorylated SAPK/JNK responded similarly to EL-EPS.

Strengths of the Study

To the best of our knowledge, the present study is the first to examine and compare the intra- and extracellular metabolome of myotubes after EL-EPS. Importantly, although we did not use tracers, we included analysis of the cell-free media controls with or without EL-EPS to the present study. This enabled us to roughly estimate whether the metabolites were taken up to the cells or released into the media. Moreover, these cell-free controls also validated the experiments showing that the effects of EL-EPS on media metabolite levels were most probably through myotube contractions and not through unspecific effects of EL-EPS on media. In addition, the number of studies analyzing the effects of glucose availability on the myotube metabolism after EL-EPS is surprisingly small (21), although the nutritional status is known to regulate many intracellular processes even under nonexercising conditions (69). That said, the limitation of the existing literature is that not all studies have clearly reported the medium glucose content (e.g., only a few of the EL-EPS articles reviewed by Nikolić et al. (16)] although it is highly recommended in other in vitro studies (17, 70). Overall, our study improves understanding of the role of nutrient availability on the metabolic changes induced by EL-EPS.

Limitations of the Study

In the present study, the selected metabolomics platform (¹H NMR) provided an introductory insight into the differences in the C2C12 myotube and media metabolites, whereas future studies using mass spectrometry (MS)-based metabolomics would be beneficial for the analysis of lower abundance metabolites. Indeed, combination of the NMR and MS platforms in metabolomics research has been recommended (71). Moreover, studies using dynamic turnover of metabolomics [i.e., fluxomics (72)] are warranted to investigate the flux of the metabolites in vitro and in vivo. We also acknowledge that we detected some minor differences between LG and HG DMEMs in some metabolites in addition to glucose (Supplemental Table S1), but unlike myo-inositol that is derived from glucose, those are likely explained by batch-to-batch differences. Finally, time series data collection after different modes of EL-EPS in upcoming studies would provide a broader understanding of the metabolite characteristics in recovery phase both in the cells and in the media.

Conclusions

By using the C2C12 myotubes, we found that EL-EPS enhanced energy source utilization as well as production and secretion of lactate, acetate, and BCFAs (see summary in Fig. 8). Many of the EL-EPS induced changes in the myotube and/or media metabolites, such as contents of lactate, acetate, BCFAs, and TCA cycle intermediates, were affected by the glucose availability. This is possibly at least in part because low-glucose condition almost fully depleted glycolytic C2C12 cells from glucose in 24 h. Thus, we recommended to consider the effect of nutrition and the choice of media in future EL-EPS studies. Lastly, the novel increase in BCFAs with EL-EPS leads to the enticing notion that these metabolites may act as exerkines warranting more research on their physiological significance and regulation by in vivo exercise.

Figure 8.

Summary of the effects of exercise-like electrical pulse stimulation (EL-EPS) on C2C12 myotube metabolism. The application of EL-EPS increased ATP production pathways, including glycolysis and lipolysis. The resulting pyruvate and acetyl-CoA were converted to lactate by lactate dehydrogenase (LDH) and to acetate possibly through reactive oxygen species (ROS) and pyruvate dehydrogenase (PDH). A buildup of several TCA intermediates suggests an overall reduction in oxidative metabolism. This is also consistent with the observed reduction in fatty acid oxidation suggested by the reduced oleate metabolism. The intermediates of the branched-chain amino acid (BCAA) catabolism include 2-methylbutyrate, isobutyrate, and isovalerate that are in part produced by the acyl-CoA thioesterases (ACOTs), possibly ACOT9. These BCAA breakdown products belong to the branched-chain fatty acids (BCFAs) and they can be directed to the TCA cycle, production of ketone bodies (e.g., 3-hydroxybutyrate) or as we observed they might also be released out of the cells. Blue = decreased, red = increased, bold black = unchanged, and black = undetected content.

SUPPLEMENTAL DATA

Supplemental Figs. S1–S5 and Supplemental Tables S1–S4: https://doi.org/10.6084/m9.figshare.14376413.v1.

GRANTS

This work was funded by the Academy of Finland Grant 298875 (to H. Kainulainen), 308042 (to S. Pekkala), 323435 (to P. Permi), and 275922 (to J. J. Hulmi). T. M. O’Connell is supported by grants from National Institutes of Health, National Institute of Arthritis and Musculoskeletal, and Skin Diseases (P01AG039355 and P30AR072581) as well as the Additional Ventures, Single Ventricle Research Fund.

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

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