Journal Logo


Substrate-Specific Respiration of Isolated Skeletal Muscle Mitochondria after 1 h of Moderate Cycling in Sedentary Adults


Author Information
Medicine & Science in Sports & Exercise: July 2021 - Volume 53 - Issue 7 - p 1375-1384
doi: 10.1249/MSS.0000000000002615
  • Free


Skeletal muscle mitochondria are a primary site for oxidative metabolism to meet energetic demands of exercise. Lipids are predominate fuel source at low to moderate exercise intensity (1,2). Aerobic exercise stimulates several regulatory points of lipid metabolism, including greater citrate synthase activity (3,4) and translocation of membrane lipid transporters (5) to increase lipid oxidation (6). Mitochondrial lipid oxidation is the net contribution of regulation intrinsic to mitochondria, such as control of oxidative phosphorylation and electron transfer system, and those extrinsic to mitochondria, including lipid transport. Lipid oxidation in mitochondria relies on electron input through electron transfer flavoprotein (ETF), which is distinct from respiration supported by NADH (N-linked, complex I) and succinate (S-linked, complex II). It remains unclear if prior aerobic exercise stimulates the intrinsic ability of mitochondria to oxidize lipids through ETF and if the response differs from nonlipid substrates.

ETF is an inner mitochondrial membrane protein complex with a primary role in lipid oxidation (7), and it is a heterodimer of α and β subunits. ETF transfers reducing equivalents generated from dehydrogenases in β-oxidation of fatty acids and amino acid oxidation into the electron transfer system (8). ETF has lower capacity for electron transfer than other complexes and produces reactive oxygen species (9). There is potential for ETF regulation through posttranslational modifications, including trimethylation of ETF-β (10) and removal of flavin from ETF (11), with each decreasing ETF activity. We reported that aerobic training in mice increased mitochondrial lipid oxidation capacity alongside increased protein abundance for ETF (12). Despite the role of ETF in lipid oxidation and ability for in vivo regulation, little is known about aerobic exercise on skeletal muscle ETF.

Aerobic exercise induces dynamic changes in oxidative metabolism that can last for several hours after exercise. Respiration rates vary between substrates and must be considered for the regulation of mitochondrial oxidation after exercise. Within seconds of starting exercise, aerobic metabolism increases to sustain ATP turnover (13). Pyruvate dehydrogenase activation increases within 30 s to promote pyruvate flux through mitochondria (14). Short bouts of high-intensity exercise (130% V˙O2peak) increased the respiration of permeabilized fibers that remained increased at 110 min after exercise (15). Higher-intensity exercise, compared with moderate intensity, is associated with exhaustion and less reliance on lipid oxidation. Time trials of 5 km at over 90% peak aerobic capacity (V˙O2peak) lowered the respiration of complex I and complex II substrates in trained adults (16). Trewin et al. (17,18) reported lower CII respiration after high-intensity intervals in obese adults. Oxidative phosphorylation of combined complex I and complex II substrates in isolated mitochondria or permeabilized fibers did not change up to 3 h after cycling at 75% V˙O2peak (4,19). Using a complex I substrate, Madsen et al. (20) reported greater respiratory control (greater oxidative phosphorylation relative to uncoupled respiration) at 30 and 60 min after exercise. Using isolated mitochondria respiring lipid substrates, the high-intensity exercise used by Tongokoni et al. (15) increased ATP production, which increased P:O. It remains unclear if respiration that is specific to mitochondria changes in the several hours after aerobic exercise and varies between substrates. Understanding intrinsic mitochondrial lipid oxidation requires specific substrates for ETF compared with complex I and complex II.

Our purpose was to determine whether a single session of moderate-intensity aerobic exercise would increase the respiration of isolated mitochondria, without influence of membrane transport or other factors extrinsic to mitochondria, in the several hours after exercise and if changes were specific to a substrate. We were further interested in the regulation of ETF as potential mediator of lipid respiration. We hypothesized that compared with a rested state, a single session of moderate-intensity cycling would increase mitochondrial respiration specific for lipids. We considered that a possible mechanism would be lower inhibitory posttranslational modifications to ETF. To test the hypothesis, we collected muscle biopsies at rest and 15 min after a single session of cycling in sedentary adults. Primary outcomes were high-resolution respirometry of isolated mitochondria for lipid and nonlipid substrates with measures of ETF protein abundance and modifications.


Overall protocol

The protocol was approved by the Institutional Review Board at Oregon State University (IRB no. 7605) and registered at (no. NCT02987491). The current evaluation of mitochondria was from baseline biopsies collected during a larger project investigating glucose and lipid metabolism after a single session of aerobic exercise (21). All study procedures were performed at the Samaritan Athletic Medical Center on the campus of Oregon State University. The study included four visits in total: screening, peak aerobic capacity (V˙O2peak), and two metabolic study days (rest and exercise). During the screening visit, a study team member explained all procedures to the participants before obtaining written consent. A fasting blood draw was collected for clinical blood panel (analyzed at Samaritan Regional Medical Center). Anthropometrics were collected, including height, weight, and body composition via dual-energy x-ray absorptiometry (Horizon; Hologic, Marlborough, MA). Females were not pregnant as verified using point-of-care testing on urine. At least 7 d later, eligible participants returned for a graded exercise test on a cycle ergometer with 12-lead electrocardiogram. Visits 3 and 4 were metabolic testing days performed in randomized order for either resting or exercise and separated by at least 7 d.


Participants were sedentary and generally free of major medical conditions (Table 1). Inclusion criteria included males and females 18–45 yr old and body mass index between 18 and 26 kg·m−2. Exclusionary criteria included structured physical activity (defined as greater than 60 min·wk−1 for more than 6 months), change in body weight (greater than 2 kg in previous 6 months), hyperglycemia (fasting glucose >126 mg·dL−1), cancer, heart disease, pregnancy, untreated hypo- or hyperthyroid, smoking, and allergy to lidocaine. Exclusionary medications included insulin, β-blockers, metformin, thiazolidinediones, statins, and chronic nonsteroidal anti-inflammatory medications. Females were studied during early follicular phase of menstrual cycle for each metabolic study day.

TABLE 1 - Participant characteristics.
Sex 4 M/11F
Age (yr) 28 + 7 (19–44)
Height (m) 1.7 + 0.1 (1.5–1.8)
Weight (kg) 61.6 + 10.3 (49.0–86.3)
Body mass index (kg·m−2) 22.2 + 2.1 (18.8–26.1)
Body fat (%) 27 + 5.3 (13–34)
Fat-free mass (kg) 45.0 + 8.2 (34.3–62.9)
V˙O2peak (L·min−1) 2.01 + 0.4 (1.44–2.82)
V˙O2peak (mL·kg−1 body weight·min−1) 32.9 + 4.5 (23.6–40)
Data are presented as mean + SD (range).


Participants arrived without eating for prior 2 h then were fitted with 12-lead electrocardiogram. The V˙O2peak test was performed on an electronically braked cycle ergometer (Corival; Lode, Groningen, The Netherlands) with indirect calorimetry (Parvo Medics TrueOne 2400, Sandy, UT). The protocol was warm-up for 2 min at 50 W, and then intensity increased each minute by 25 W for males and 15 W for females. RPE (6–20 on Borg scale) and blood pressure were collected every 3 min. V˙O2peak was defined as peak oxygen consumption during 30-s averaging achieved with 1) failure to maintain pedaling cadence of at least 60 rpm, 2) reaching approximately 10% of age-predicted maximal heart, and 3) respiratory exchange ratio greater than 1.1.

Biopsy collection

Skeletal muscle biopsies were collected from vastus lateralis on two visits separated by at least 7 d. Participants avoided exercise for 3 d before metabolic study days and recorded all food intake for 24 h preceding the study visit and then replicated dietary intake for the second study visit. Participants arrived after an overnight fast (10 h, water okay) at approximately 0700 h. Biopsies were collected using Bergstrom technique modified with suction. Sites were on lateral aspect of thigh at approximately 4 cm proximal from knee. Local anesthetic was 2% lidocaine. For resting conditions, participants lay in bed for 1 h. For exercise conditions, participants cycled for 1 h on an electronically braked ergometer. Workload was adjusted to maintain 65% V˙O2peak, verified using indirect calorimetry during the first ~20 min of exercise. The participant returned to bed, and a biopsy was collected at 15 min postexercise or rest. Biopsy samples were sectioned (while chilled on ice) for respiration, and the remaining sample was frozen in liquid nitrogen then stored −80°C.

Mitochondrial respiration

Mitochondria were isolated from fresh biopsy samples then analyzed using high-resolution respirometry (Oxygraph O2K; Oroboros Instruments, Innsbruck, Austria). Approximately 100 mg of tissue was minced and incubated in buffer A (100 mM KCl, 50 mM Tris, 5 mM MgCl2, 1.8 mM ATP, and 1 mM EDTA, pH 7.2) with protease (Subtilisin A, Sigma P5380) for 7 min on ice and then homogenized in glass-on-glass homogenizers with 0.3-mm spacing between mortar and pestle for 10 min at 150 rpm. Samples were centrifuged for 5 min at 750g and 4°C, and then the supernatant was removed and centrifuged for 5 min at 10,000g and 4°C to pellet the mitochondria. The supernatant was discarded, and the mitochondrial pellet was washed with buffer A and centrifuged 5 min at 9000g and 4°C. The final mitochondrial-enriched pellet was resuspended in 1:4.2 (wt/vol) buffer B (180 mM sucrose, 35 mM KH2PO4, 10 mM Mg-acetate, 5 mM EDTA, pH 7.5).

High-resolution respirometry was performed using Oxygraph-2k units (Oroboros Instruments) with MiR05 respiration buffer (0.5 mM EGTA, 3 mM MgCl2-6H2O, 60 mM lactobionic acid, 20 mM taurine, 10 mM KH2PO4, 20 mM HEPES, 110 mM sucrose, and 1 g·L−1 bovine serum albumin). Instrument settings were 37°C, stirring at 750 RPM and 2-s data averaging (Datlab 6, Oroboros Instruments). The oxygen concentration in chambers was maintained between 50 and 200 nmol·mL−1. We conducted two independent protocols using lipid substrate for ETF (palmitoylcarnitine and malate) or nonlipid substrates (glutamate, malate, and succinate) to determine rates of oxygen consumption (JO2). Hydrogen peroxide (JH2O2) emission was measured simultaneously using 10 μM Amplex red, 5 U·mL−1 superoxide dismutase, and 1 U·mL−1 horseradish peroxidase (calibrated with H2O2 injections) as previously described and with a few variations (22). ADP-limiting conditions generate reverse electron flow and greater leak of electrons to produce reactive oxygen species (indicating capacity for production), whereas ADP saturating conditions stimulate forward flow and less leak to oxygen (more representative of oxidative phosphorylation), particularly through complex I (23). Each protocol was performed in duplicate chambers (A and B) on one machine, and the rate of oxygen consumed (pmol O2·mL−1·s−1) at each point was calculated as the average values from the two chambers. The protein concentration of the mitochondrial preparation was measured by Pierce BCA assay to calculate JO2 relative to protein content (pmol O2·μg−1 mitochondrial protein·s−1).

We determined oxidative phosphorylation (P), leak (L), and noncoupled state respiration (E) for lipid and nonlipid protocols. The titration sequence for lipid protocol was 90 μL of mitochondrial suspension, 25 μM palmitoylcarnitine + 2 mM malate, ADP pulse (113.6 nmol), 2.5 mM ADP (P, F-linked), 20 μM cytochrome c for membrane integrity test (PCytC), 2 μg·mL−1 oligomycin (LO, F-linked), sequential additions of 0.5 μM FCCP to plateau (E, F-linked), and 2.5 μM antimycin A (residual oxygen consumption). The titration sequence for nonlipid protocol was 90 μL of mitochondrial suspension, 10 mM glutamate + 2 mM malate, ADP pulse (113.6 nmol), 2.5 mM ADP (P, N-linked), 10 mM succinate (P, NS-linked), 20 μM cytochrome c (PCytc, NS-linked), 2 μg·mL−1 oligomycin (LO, NS-linked), sequential additions of 0.5 μM FCCP to plateau (E, NS-linked), 0.5 μM rotenone (E, S-linked), and 2.5 mM antimycin A. We determined P:O ratio as oxygen consumption during the subsaturating ADP pulse. The P:O ratio assumes ADP consumption, and ATP production occurs in 1:1 molar ratio. Coupled respiration was calculated by subtracting leak respiration from maximal oxidative phosphorylation (LO minus PCytC each for F- and NS-linked protocols). Oligomycin-induced leak drives high membrane potential without ATP production and is an upper limit of nonphosphorylating leak respiration. Subtracting leak calculates respiration that is fully coupled to ATP production. Leak respiration varies between respiratory states and was also determined for each substrate. Electron leak to O2 was calculated as the rate of H2O2 emission divided by two times the simultaneous rate of O2 consumption then multiplied by 100 (24).

Skeletal muscle protein abundance

Frozen muscle samples were powdered and homogenized (~40 mg) for immunoblotting. Homogenates were rotated for 20 min at 4°C then centrifuged at 10,000g for 20 min and supernatant collected. Protein concentration was determined using bicinchoninic acid assay (ThermoFisher Scientific, Waltham, MA). Samples were diluted in Laemmli buffer, and approximately 30 μg of protein was resolved on 10%–15% Bis–Tris gels then transferred to nitrocellulose membranes using semidry transfer (TransBlot Turbo; Bio-Rad, Hercules, CA). A control sample was loaded at the beginning and end of each gel to serve as an internal control, and the average intensity was used to normalize band intensities between gels. Ponceau staining of membranes was performed to verify equal loading and transfer of protein to the nitrocellulose membrane. Membranes were blocked in 5% bovine serum albumin (BSA) in Tris-buffered saline +1%Tween (TBST). Primary antibodies were diluted in 5% BSA–TBST or 5% nonfat dry milk–TBST, and membranes were incubated overnight at 4°C. Secondary antibodies were diluted in 5% BSA–TBST or 5% nonfat dry milk–TBST, and membranes were incubated at room temperature for 1 h. Images were generated using infrared detection (Odyssey; Licor, Lincoln, NE) and analyzed using ImageView software (Licor). Primary antibodies were all diluted 1:1000 for ETF-α (Abcam no. 110316), ETF-β (Abcam no. 240593), trimethylated ETF-β (Abcam no. 76118), total PDH E1α (Abcam no. 168379), phosphorylated PDH E1α at S293 (Abcam no. 177461), and β-HAD (ThermoFisher PAS-28203).


The study was a repeated crossover design. Primary outcomes were compared using a two-tailed paired t-test. Statistical significance was set at α = 0.05. Power calculation for group sizes was based on previous reports. The respiration of isolated mitochondria using glutamate–malate–succinate substrates in similar age-group (n = 34, 18–30 yr, untrained) revealed a mean of 5.1 and an SD of 1.6 pmol·μg−1 mitochondrial protein·s−1 (25). Using these estimates and α = 0.05, a sample size of 15 would have a power (1 − β) of 0.8 or 0.5 and detect differences of 1.6 and 1.2 (~31% and 23%, respectively) with effect sizes of 1.0 or 0.75. We reasoned that detecting changes larger than 20% would be physiologically meaningful on mitochondrial respiration after exercise, beyond technical and biological variability (26). Data are presented as mean and SD as recommended in the field (27), with individual data displayed when possible.


Mitochondrial respiration

We determined the respiration of mitochondrial-enriched samples collected from skeletal muscle at rest or 15 min after aerobic exercise. Citrate synthase activity is a marker of mitochondria abundance and was not different between rest and exercise (data not shown), indicating consistent isolation yield between study days. Respiration data are normalized to the protein concentration of the mitochondria pellet. We included males and females but were not powered to detect sex differences. We used separate titration protocols of palmitoylcarnitine (for F-linked respiration through ETF) versus glutamate and succinate (for N- and S-linked respiration, respectively). Membrane integrity of the mitochondrial preparation was verified by minimal change in respiration after the addition of cytochrome c (<3% change and P > 0.05 compared with prior steady-state respiration). Nonmitochondrial respiration, determined using the addition of antimycin A (complex III inhibitor), was less than 5% of maximal respiration and higher after exercise for nonlipids (P = 0.06 for rest vs postexercise of 0.11 and 0.14 pmol·μg−1 mitochondrial protein·s−1). One data point was exceptionally high for ADP-stimulated respiration. We include the data point in the final analysis and graphs because 1) there was no clear reason to exclude for technical reasons and 2) it did not influence conclusions.

We were first interested in the effects of exercise compared with rest on maximal oxidative phosphorylation capacity of specific substrates. During the lipid respiration protocol (Fig. 1), postexercise oxidative phosphorylation tended to be higher compared with rest (+12%, P = 0.09). Compared with rest, exercise did not alter respiration for the leak state (an indication of membrane permeability) when measured for individual substrates or after oligomycin. Noncoupled respiration (an indication of electron transfer system) was also not different. We next determined respiration using glutamate–malate–succinate for N- and S-linked respiration (complex I and complex II, respectively; Fig. 2). Again, postexercise had tendency for greater oxidative phosphorylation when measured during combined input through N + S substrates (+14%, P = 0.09). Inhibiting complex I input, using rotenone, and respiring via succinate eliminated the small exercise effects. Like lipid-supported respiration, the succinate-supported leak and noncoupled respiration states were not altered by exercise.

F-linked respiration after aerobic exercise. Respiration of isolated mitochondria collected from vastus lateralis biopsies at rest or 15 min after aerobic exercise (60 min at 65% V˙O2peak). There were no major changes in respiration for any respiratory state with tendency for greater oxidative phosphorylation. Labels above the graph indicate respiratory states for leak (L), oxidative phosphorylation (P), and electron transfer system (E) for the lipid (F-linked) respiration of mitochondria (mitos), palmitoyl carnitine + malate (PC + M), saturated ADP (ADP), cytochrome c for membrane integrity, oligomycin (LO), and uncoupling via FCCP and antimycin A for residual oxygen consumption (Rox). Data are displayed as mean and SD with individual data points (n = 15 males and females). Data were compared using paired t-test between rest and exercise within each respiratory state.
N- and S-linked respiration after aerobic exercise. Respiration of isolated mitochondria collected from vastus lateralis biopsies at rest or 15 min after aerobic exercise (60 min at 65% V˙O2peak). There were no major changes in respiration for any respiratory state with tendency for greater oxidative phosphorylation. Labels above the graph indicate respiratory states for leak (L), oxidative phosphorylation (P), and electron transfer system (E) for glutamate + malate (N-linked) and succinate (S-linked) respiration of mitochondria (mitos), glutamate + malate (GM), ADP, cytochrome c for membrane integrity, rotenone for complex I inhibition, oligomycin (LO), electron transfer system (E) during uncoupling via FCCP, and antimycin A for residual oxygen consumption (Rox). Data are displayed as mean and SD with individual data points (n = 15 males and females). Data were compared using paired t-test between rest and exercise within each respiratory state.

We considered the coupled respiration of each substrate after aerobic exercise. Coupled oxidative phosphorylation is oxidative phosphorylation minus leak respiration. Exercise tended to increase coupled oxidative phosphorylation for F-linked (+13%, P = 0.08), N + S-linked (14%, P = 0.09), and N-linked (+17%, P = 0.08) (Fig. 3A). P:O ratio is an index of phosphorylation efficiency and was not changed after exercise for either substrate (Fig. 3B). Our measured P:O for lipids was ~3.5 and for glutamate–malate was ~2.5, whereas the P:O is approximately 2.8 for lipids and 3.0 for carbohydrates (28). The discrepancy from estimates is not clear and may be related to isolation methods that disrupt the mitochondrial reticulum. We used aliquots of ADP from a single batch to compare P:O between study days. Increased coupled respiration, and not the leak state (for individual substrates or oligomycin), contributed to the ~10% increase in oxidative phosphorylation after aerobic exercise.

Coupled respiration after aerobic exercise. Coupled respiration determined as oxidative phosphorylation minus leak respiration for each respiratory substrate (A). P:O was calculated as oxygen consumption during a subsaturating pulse of ADP and was not changed after exercise (B). Data are displayed as mean and SD with individual data points (n = 15 males and females). Data were compared using paired t-test between rest and exercise within each respiratory state.

Electron leak to H2O2

We measured H2O2 emission (JH2O2) with JO2 across respiratory state and substrates to better understand the production of reactive oxygen species from mitochondria after aerobic exercise (Fig. 4). Exercise did not alter H2O2 emission per mitochondrial protein in any state (Fig. 4A). Electron leak to H2O2 accounts for simultaneous oxygen consumption. Exercise induced a small increase of electron leak to H2O2 during the leak respiration of PC + M but not during phosphorylating conditions for any substrates (Fig. 4C and D). Electron leak to H2O2 during the leak state (ADP limited) was about 5.5% for glutamate–malate–succinate and about 2% for palmitoylcarnitine. Electron leak to H2O2 during ADP-stimulated oxidative phosphorylation was ~0.2% for palmitoylcarnitine and about 0.05% for glutamate–malate–succinate. The differences in H2O2 emission between substrates were larger than exercise effects.

H2O2 emission and electron leak after aerobic exercise. H2O2 emission was measured during respiratory protocols from isolated mitochondria collected from vastus lateralis biopsies at rest or 15 min after aerobic exercise (60 min at 65% V˙O2peak). Prior exercise did not change H2O2 emissions for F-linked (A) or N(S)-linked substrates (B). There is a small effect of exercise on electron leak to H2O2 during the leak respiration of F-linked substrates (C) with no changes in N(S) substrates (D). Note the broken y-axis scales due to the wide range of electron leak during leak (L) vs oxidative phosphorylation (P). Data are displayed as mean and SD with individual data points (n = 15 males and females). Data were compared using paired t-test between rest and exercise within each respiratory state.

Protein abundance

We determined skeletal muscle ETF protein abundance and trimethylation (inhibitory modification). A single bout of aerobic exercise did not change the protein abundance of ETF subunits α or β (Fig. 5A and B). Aerobic exercise did not alter ETF-β methylation (Fig. 5C) or β-HAD abundance, an upstream dehydrogenase (Fig. 5D). We did not detect the activation of PDH via phosphorylation at 15 min after exercise (Fig. 5F and G). A single session of aerobic exercise does not appear to change protein abundance for ETF or methylation of β subunit.

Protein abundance after aerobic exercise. Protein abundance via Western blotting of biopsy samples collected from vastus lateralis during rest or 15 min after aerobic exercise (60 min at 65% V˙O2peak). There were no changes in protein abundance after exercise for ETF-α or ETF-β subunits, nor in the inhibitory modification of the trimethylation of ETF-β (A–C). β-HAD abundance was also not different (D). There were no differences between rest and 15 min postexercise for PDH E1α or phosphorylated PDH E1α (F and G). Representative images are displayed in panels E and H. Data are displayed as mean and SD with individual data points (n = 15 males and females). Data were compared using paired t-test between rest and exercise within each blot.


We investigated the respiration of isolated mitochondria in the several hours after a single session of aerobic exercise in sedentary adults. We used multiple substrates for respiration with a focus on lipid oxidation and considered ETF as a potential regulatory point of mitochondrial metabolism in response to aerobic exercise. A single session of moderate aerobic exercise had a modest effect to increase maximal respiratory capacity for lipid substrates (ETF-supported respiration) or N- and S-linked substrates (which do not rely on ETF). Effect sizes for exercise on coupled respiration were approximately 0.3 for F-linked through 0.5 for N(S)-linked (~13%). H2O2 emission also had minor increases for only lipid substrates. There were no changes to the ETF protein abundance or the methylation of ETF-β. The lower respiration of lipid substrates than complex I or II substrates is consistent with lipid respiration being limited at or upstream of ETF. Any stimulatory effect of aerobic exercise on intrinsically controlled mitochondrial respiration was small, relative to respiratory capacity, and consistent across several substrates.

Our findings provide several insights into mitochondrial lipid oxidation after aerobic exercise. First, we measured mitochondrial metabolism using separate substrates. In our study of 15 sedentary adults, a single session of exercise induced modest increases in maximal oxidative phosphorylation capacity for palmitoylcarnitine, a predominate intramuscular lipid species. Our untrained participants were free of major disease but were on the lower end of fitness, which associates with low mitochondrial abundance and ATP production (29). The average V˙O2peak of the current participants was in the lower third percentile for females and 10th percentile for males, indicating very low fitness capacity (30). Previous studies indicate that prior exercise alters mitochondrial response regardless of training status of the person. For example, aerobic exercise lowered respiration when measured using combinations of lipids with other substrates in either well-trained cyclists (31) or adults with obesity (18). Lipid-only respiration did not change after exercise in well-trained cyclists consuming either high-fat or high-protein diet for 5 d (effect sizes, 0.78 and 0.17, respectively), suggesting that short-term changes in diet may also not affect lipid respiration to aerobic exercise (31). Effect sizes of prior exercise reported by Layec et al. (16) were ~5 for lower complex I and complex II respiration but ~1.5 for complex I. We measured lipids separate from complex I and complex II. Measuring individual substrates is necessary because oxidative metabolism shifts from carbohydrate oxidation at higher intensities to lipid oxidation during longer bouts at moderate intensity. For example, 120 min of cycling, but not 30 min, increased palmitate oxidation rates in recreationally active adults (6). We do not anticipate that higher intensity would alter results as previous investigations did not reveal any changes to maximal ATP production rate for lipid substrates after exercise at 130% V˙O2peak (15). Previous work indicated that 1 h of cycling improved ADP sensitivity and lowered the Km for complex I (32). V˙O2 during exercise is a function of intramuscular ADP concentration that changes through onset of exercise to steady state with the rapid activation of mitochondria (33). We measured maximal respiration during ADP saturated conditions and cannot exclude changes in substrate sensitivity.

Second, the single session of exercise did not change ETF protein abundance. We considered ETF as a regulator of lipid oxidation because ETF is the final transfer complex for reducing equivalents from FADH2 to the electron transfer system (34). Patients with impaired ETF activity have low lipid oxidation, accumulation of intramuscular lipids, and exercise intolerance (35,36). Long-term changes in ETF protein abundance, such as with training, may contribute to exercise improvements in mitochondrial lipid oxidation (12). Mitochondrial proteins have fractional synthesis rates of 0.1% to 0.2% per hour at rest and postexercise (25,37,38), which, when countered with protein degradation, are likely not high enough to detect differences in protein abundance after 1 h. ETF has several regulatory points, including inhibition by trimethylation (10). We did not detect changes in ETF methylation status with exercise, which aligns with no major changes in respiration. Lipid respiration (during oxidative phosphorylation, P) was ~35% lower for lipid substrates than for complex I. Although required for lipid respiration, complex I has excess capacity for electron transfer compared with ETF and does not appear to limit lipid respiration (12,22,39).

More upstream to ETF, the delivery of fatty acids can stimulate fatty acid oxidation after aerobic exercise (5). Throughout 120 min of cycling, exercise stimulated progressive increases in the abundance of the lipid transporter CPT1 alongside increases in mitochondrial lipid oxidation (6). Higher exercise intensities decrease lipid oxidation, yet maintaining plasma free fatty acids attenuates the drop in lipid oxidation at higher exercise intensities (40). High-intensity exercise modifies CPT1 flux with greater inhibition and blunted increases in lipid oxidation (41). These studies emphasize the importance of fatty acid delivery to regulate muscle fat oxidation to changing exercise intensities. More downstream, the activity of β-HAD did not change after repeated bouts of a higher-intensity cycling (6 min at 91% V˙O2peak) (42). Intramuscular lipids also support lipid oxidation in addition to delivery from circulation (43). We focused on intrinsic mitochondrial metabolism through ETF as a potential regulation site of fat oxidation. Our current data indicate that 60 min of moderate-intensity exercise did not alter ETF protein abundance.

Third, we report exercise had small effect for greater H2O2 emission for lipids after exercise. Electron leak to H2O2 was about three to four times greater during oxidative phosphorylation of fatty acids than for nonlipid substrates for complex I or complex II. Flux through ETF, more so than complex I or complex II, produces reactive oxygen species during nonphosphorylating fatty acid respiration (9). Trewin et al. (17) reported that H2O2 emissions during the leak respiration of complex I and complex II substrates decreased right after moderate, work-matched high-intensity intervals, and sprint intervals but increased for oxidative phosphorylation at 3 h after exercise. Our study adds new information of H2O2 emission through separate lipid and nonlipid substrates, which have different rates of electron leak (44). Fatty acid–induced reactive oxygen species are implicated in the development of metabolic disease associated with sedentary lifestyles such as obesity (45). Our current data in sedentary males and females demonstrate a minor increase of electron leak to H2O2 during the respiration of ETF substrates, but not complex I or complex II substrates, from resting to postexercise.

We performed respiration on mitochondrial-enriched samples to determine the intrinsic regulation of respiration. We did not detect the activation of PDH at 15 min postexercise, which activates within seconds of exercise and deactivates within minutes of rest (14). We are therefore not measuring acute exercise but the lasting effects of prior exercise. Sample processing requires about 30 to 45 min, so changes to mitochondria would need to withstand isolation techniques. Our approach provided direct investigation of mitochondrial respiration without diffusion limitations and tested mitochondria from a larger muscle sample to minimize the effect of tissue heterogeneity (such as fiber type). We used palmitoylcarnitine, which avoids limitations related to carnitine transfer through mitochondrial membrane. Differences between fiber preparations and mitochondrial isolation are critical considerations as exercise alters fatty acid uptake at the cell membrane, which is removed during cell fractionation (46). The isolation procedure represents a larger fraction of mitochondria because it starts from 90 to 100 mg of muscle tissue, versus 2–3 mg for fiber preparations. Direct comparisons between respiration rates reveal similar qualitative findings such as with aging and caloric restriction (47). Also, intermyofibrillar mitochondria have higher respiratory rates compared with subsarcolemmal mitochondria (48). Our isolation includes protease treatment to liberate both populations. Extrapolating our lipid respiration rates to whole body results in approximately 0.64 g fat oxidation per kilogram lean mass per minute (assuming mitochondria at 5% cell volume (49) and wet-to-dry protein conversion of 4.35 [50]), which align with whole-body indirect calorimetry measures. We measured fat oxidation of 2.04 μmol fatty acids per kilogram lean mass per minute in these participants (data not provided), indicating good integration of the measures of isolated mitochondria to whole-body physiology. Our approach provides insight into mitochondrial physiology while attempting to minimize technical and biologic variability (26).

In conclusion, our results indicate that prior aerobic exercise induced a modest increase in coupled oxidative phosphorylation that was consistent across lipid and nonlipid substrates. Mitochondria have clear excess capacity for oxidative phosphorylation beyond lipid respiration. Integrating our results with the field indicates that increases in intrinsic mitochondrial oxidative capacity combined with greater substrate trafficking and delivery to tissue promote lipid oxidation with exercise. Next steps are to consider the interplay of substrates (such as upstream flux through β-oxidation and TCA cycle) to the electron transfer system after aerobic exercise. The role of ETF in this juncture remains to be elucidated in lipid metabolism such as with exercise training or obesity.

The authors thank the participants for their participation. They also thank the staff of Samaritan Athletic Medical Center, including clinical oversight by Drs. Nicholas Phillips, Craig Graham, and Joshua Lenhoff. They appreciate the contributions of Emily Burney and Bergen Sather for assistance during metabolic study days and Philip Batterson with the P:O calculations. Study funding was provided by KL2TR002370 awarded to S. A. N. as part of the Oregon Clinical and Translational Research Institute’s Clinical and Translational Science Award from the National Institutes of Health, the Collins Medical Trust awarded to S. A. N. and M. M. R., and the John C. Erkkila, M.D. Endowment for Health and Human Performance awarded to S. A. N. and M. M. R. H. D. S. and S. E. E. were supported by fellowships from Oregon State University. MMR was supported by K01DK103829.

The authors have no conflicts of interest to disclose and declare that study results are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation. The results of the present study do not constitute endorsement by the American College of Sports Medicine.


1. Bergman BC, Butterfield GE, Wolfel EE, Casazza GA, Lopaschuk GD, Brooks GA. Evaluation of exercise and training on muscle lipid metabolism. Am J Physiol. 1999;276(1 Pt 1):E106–17.
2. Friedlander AL, Casazza GA, Horning MA, Buddinger TF, Brooks GA. Effects of exercise intensity and training on lipid metabolism in young women. Am J Physiol. 1998;275(5):E853–63.
3. Leek BT, Mudaliar SR, Henry R, Mathieu-Costello O, Richardson RS. Effect of acute exercise on citrate synthase activity in untrained and trained human skeletal muscle. Am J Physiol Regul Integr Comp Physiol. 2001;280(2):R441–7.
4. Fernstrom M, Tonkonogi M, Sahlin K. Effects of acute and chronic endurance exercise on mitochondrial uncoupling in human skeletal muscle. J Physiol. 2004;554(Pt 3):755–63.
5. McFarlan JT, Yoshida Y, Jain SS, et al. In vivo, fatty acid translocase (CD36) critically regulates skeletal muscle fuel selection, exercise performance, and training-induced adaptation of fatty acid oxidation. J Biol Chem. 2012;287(28):23502–16.
6. Holloway GP, Bezaire V, Heigenhauser GJ, et al. Mitochondrial long chain fatty acid oxidation, fatty acid translocase/CD36 content and carnitine palmitoyltransferase I activity in human skeletal muscle during aerobic exercise. J Physiol. 2006;571(Pt 1):201–10.
7. Roberts DL, Frerman FE, Kim JJ. Three-dimensional structure of human electron transfer flavoprotein to 2.1-a resolution. Proc Natl Acad Sci U S A. 1996;93(25):14355–60.
8. Ramsay RR, Steenkamp DJ, Husain M. Reactions of electron-transfer flavoprotein and electron-transfer flavoprotein: ubiquinone oxidoreductase. Biochem J. 1987;241(3):883–92.
9. Seifert EL, Estey C, Xuan JY, Harper ME. Electron transport chain-dependent and -independent mechanisms of mitochondrial H2O2 emission during long-chain fatty acid oxidation. J Biol Chem. 2010;285(8):5748–58.
10. Rhein VF, Carroll J, He J, Ding S, Fearnley IM, Walker JE. Human METTL20 methylates lysine residues adjacent to the recognition loop of the electron transfer flavoprotein in mitochondria. J Biol Chem. 2014;289(35):24640–51.
11. Floyd BJ, Wilkerson EM, Veling MT, et al. Mitochondrial protein interaction mapping identifies regulators of respiratory chain function. Mol Cell. 2016;63(4):621–32.
12. Ehrlicher SE, Stierwalt HD, Miller BF, Newsom SA, Robinson MM. Mitochondrial adaptations to exercise do not require Bcl2-mediated autophagy but occur with BNIP3/Parkin activation. FASEB J. 2020;34(3):4602–18.
13. Bangsbo J, Krustrup P, Gonzalez-Alonso J, Saltin B. ATP production and efficiency of human skeletal muscle during intense exercise: effect of previous exercise. Am J Physiol Endocrinol Metab. 2001;280(6):E956–64.
14. Gurd BJ, Peters SJ, Heigenhauser GJ, et al. Prior heavy exercise elevates pyruvate dehydrogenase activity and muscle oxygenation and speeds O2 uptake kinetics during moderate exercise in older adults. Am J Physiol Regul Integr Comp Physiol. 2009;297(3):R877–84.
15. Tonkonogi M, Walsh B, Tiivel T, Saks V, Sahlin K. Mitochondrial function in human skeletal muscle is not impaired by high intensity exercise. Pflugers Arch. 1999;437(4):562–8.
16. Layec G, Blain GM, Rossman MJ, et al. Acute high-intensity exercise impairs skeletal muscle respiratory capacity. Med Sci Sports Exerc. 2018;50(12):2409–17.
17. Trewin AJ, Parker L, Shaw CS, et al. Acute HIIE elicits similar changes in human skeletal muscle mitochondrial H2O2 release, respiration, and cell signaling as endurance exercise even with less work. Am J Physiol Regul Integr Comp Physiol. 2018;315(5):R1003–16.
18. Trewin AJ, Levinger I, Parker L, et al. Acute exercise alters skeletal muscle mitochondrial respiration and H2O2 emission in response to hyperinsulinemic–euglycemic clamp in middle-aged obese men. PLoS One. 2017;12(11):e0188421.
19. Tonkonogi M, Harris B, Sahlin K. Mitochondrial oxidative function in human saponin-skinned muscle fibres: effects of prolonged exercise. J Physiol. 1998;510(Pt 1):279–86.
20. Madsen K, Ertbjerg P, Djurhuus MS, Pedersen PK. Calcium content and respiratory control index of skeletal muscle mitochondria during exercise and recovery. Am J Physiol. 1996;271(6 Pt 1):E1044–50.
21. Stierwalt HD, Ehrlicher SE, Robinson MM, Newsom SA. Skeletal muscle ACSL isoforms relate to measures of fat metabolism in humans. Med Sci Sports Exerc. 2021;53(3):624–32.
22. Newsom SA, Miller BF, Hamilton KL, Ehrlicher SE, Stierwalt HD, Robinson MM. Long-term rates of mitochondrial protein synthesis are increased in mouse skeletal muscle with high-fat feeding regardless of insulin-sensitizing treatment. Am J Physiol Endocrinol Metab. 2017;313(5):E552–62.
23. Lambert AJ, Brand MD. Inhibitors of the quinone-binding site allow rapid superoxide production from mitochondrial NADH:ubiquinone oxidoreductase (complex I). J Biol Chem. 2004;279(38):39414–20.
24. Daussin FN, Rasseneur L, Bouitbir J, et al. Different timing of changes in mitochondrial functions following endurance training. Med Sci Sports Exerc. 2012;44(2):217–24.
25. Robinson MM, Dasari S, Konopka AR, et al. Enhanced protein translation underlies improved metabolic and physical adaptations to different exercise training modes in young and old humans. Cell Metab. 2017;25(3):581–92.
26. Jacques M, Kuang J, Bishop DJ, et al. Mitochondrial respiration variability and simulations in human skeletal muscle: the gene SMART study. FASEB J. 2020;34(2):2978–86.
27. Curran-Everett D. Explorations in statistics: standard deviations and standard errors. Adv Physiol Educ. 2008;32(3):203–8.
28. Brooks GA. Bioenergetics of exercising humans. Compr Physiol. 2012;2(1):537–62.
29. Short KR, Bigelow ML, Kahl J, et al. Decline in skeletal muscle mitochondrial function with aging in humans. Proc Natl Acad Sci U S A. 2005;102(15):5618–23.
30. Kaminsky LA, Arena R, Myers J. Reference standards for cardiorespiratory fitness measured with cardiopulmonary exercise testing: data from the fitness registry and the importance of exercise national database. Mayo Clin Proc. 2015;90(11):1515–23.
31. Leckey JJ, Hoffman NJ, Parr EB, et al. High dietary fat intake increases fat oxidation and reduces skeletal muscle mitochondrial respiration in trained humans. FASEB J. 2018;32(6):2979–91.
32. Perry CG, Kane DA, Herbst EA, et al. Mitochondrial creatine kinase activity and phosphate shuttling are acutely regulated by exercise in human skeletal muscle. J Physiol. 2012;590(21):5475–86.
33. Wust RC, Grassi B, Hogan MC, Howlett RA, Gladden LB, Rossiter HB. Kinetic control of oxygen consumption during contractions in self-perfused skeletal muscle. J Physiol. 2011;589(Pt 16):3995–4009.
34. Crane FL, Mii S, Hauge JG, Green DE, Beinert H. On the mechanism of dehydrogenation of fatty acyl derivatives of coenzyme A. I. The general fatty acyl coenzyme A dehydrogenase. J Biol Chem. 1956;218(2):701–6.
35. Olsen RK, Olpin SE, Andresen BS, et al. ETFDH mutations as a major cause of riboflavin-responsive multiple acyl-CoA dehydrogenation deficiency. Brain. 2007;130(Pt 8):2045–54.
36. Takken T, Custers J, Visser G, Dorland L, Helders P, de Koning T. Prolonged exercise testing in two children with a mild multiple acyl-CoA-dehydrogenase deficiency. Nutr Metab (Lond). 2005;2(1):12.
37. Robinson MM, Bell C, Peelor FF 3rd, Miller BF. β-Adrenergic receptor blockade blunts post-exercise skeletal muscle mitochondrial protein synthesis rates in humans. Am J Physiol Regul Integr Comp Physiol. 2011;301(2):R327–34.
38. Wilkinson SB, Phillips SM, Atherton PJ, et al. Differential effects of resistance and endurance exercise in the fed state on signalling molecule phosphorylation and protein synthesis in human muscle. J Physiol. 2008;586(Pt 15):3701–17.
39. Jacobs RA, Fluck D, Bonne TC, et al. Improvements in exercise performance with high-intensity interval training coincide with an increase in skeletal muscle mitochondrial content and function. J Appl Physiol. 2013;115(6):785–93.
40. Romijn JA, Coyle EF, Sidossis LS, Zhang XJ, Wolfe RR. Relationship between fatty acid delivery and fatty acid oxidation during strenuous exercise. J Appl Physiol (1985). 1995;79(6):1939–45.
41. Petrick HL, Holloway GP. High intensity exercise inhibits carnitine palmitoyltransferase-I sensitivity to L-carnitine. Biochem J. 2019;476(3):547–58.
42. Green HJ, Bombardier EB, Duhamel TA, Holloway GP, Tupling AR, Ouyang J. Acute responses in muscle mitochondrial and cytosolic enzyme activities during heavy intermittent exercise. J Appl Physiol. 2008;104(4):931–7.
43. Romijn JA, Coyle EF, Sidossis LS, et al. Regulation of endogenous fat and carbohydrate metabolism in relation to exercise intensity and duration. Am J Physiol. 1993;265(3 Pt 1):E380–91.
44. Robinson MM, Sather BK, Burney ER, et al. Robust intrinsic differences in mitochondrial respiration and H2O2 emission between L6 and C2C12 cells. Am J Physiol Cell Physiol. 2019;317(2):C339–47.
45. Konopka AR, Asante A, Lanza IR, et al. Defects in mitochondrial efficiency and H2O2 emissions in obese women are restored to a lean phenotype with aerobic exercise training. Diabetes. 2015;64(6):2104–15.
46. Picard M, Taivassalo T, Ritchie D, et al. Mitochondrial structure and function are disrupted by standard isolation methods. PLoS One. 2011;6(3):e18317.
47. Lanza IR, Zabielski P, Klaus KA, et al. Chronic caloric restriction preserves mitochondrial function in senescence without increasing mitochondrial biogenesis. Cell Metab. 2012;16(6):777–88.
48. Palmer JW, Tandler B, Hoppel CL. Biochemical properties of subsarcolemmal and interfibrillar mitochondria isolated from rat cardiac muscle. J Biol Chem. 1977;252(23):8731–9.
49. Jacobs RA, Lundby AK, Fenk S, et al. Twenty-eight days of exposure to 3454 m increases mitochondrial volume density in human skeletal muscle. J Physiol. 2016;594(5):1151–66.
50. Areta JL, Hopkins WG. Skeletal muscle glycogen content at rest and during endurance exercise in humans: a meta-analysis. Sports Med. 2018;48(9):2091–102.


Copyright © 2021 by the American College of Sports Medicine