Utilization of lactic acid in human myotubes and interplay with glucose and fatty acid metabolism

Once assumed only to be a waste product of anaerobe glycolytic activity, lactate is now recognized as an energy source in skeletal muscles. While lactate metabolism has been extensively studied in vivo, underlying cellular processes are poorly described. This study aimed to examine lactate metabolism in cultured human myotubes and to investigate effects of lactate exposure on metabolism of oleic acid and glucose. Lactic acid, fatty acid and glucose metabolism were studied in myotubes using [14C(U)]lactic acid, [14C]oleic acid and [14C(U)]glucose, respectively. Myotubes expressed both the MCT1, MCT2, MCT3 and MCT4 lactate transporters, and lactic acid was found to be a substrate for both glycogen synthesis and lipid storage. Pyruvate and palmitic acid inhibited lactic acid oxidation, whilst glucose and α-cyano-4-hydroxycinnamic acid inhibited lactic acid uptake. Acute addition of lactic acid inhibited glucose and oleic acid oxidation, whereas oleic acid uptake was increased. Pretreatment with lactic acid for 24 h did not affect glucose or oleic acid metabolism. By replacing glucose with lactic acid during the whole culturing period, glucose uptake and oxidation were increased by 2.8-fold and 3-fold, respectively, and oleic acid oxidation was increased 1.4-fold. Thus, lactic acid has an important role in energy metabolism of human myotubes.

Lactate is produced from glucose through glycolysis and the conversion of pyruvate by lactate dehydrogenase (LDH). Once considered a waste product of anaerobic glycolysis, lactate is now recognized to be formed continuously also under fully aerobic conditions 1 . The main function of lactate is to serve as a precursor of hepatic gluconeogenesis, but also as an energy substrate for aerobic oxidation via the citric acid cycle (TCA cycle) in many peripheral tissues. In a recently published systematic analysis of the fluxes of circulating metabolites in mice, lactate was revealed as a primary source of carbon for the TCA cycle, demonstrating that the contribution of glucose to the TCA cycle in majority of tissues (except brain) occurs through circulating lactate 2 . The major organ of lactate production and utilization is skeletal muscle, especially during exercise. While substantial amount of research has been devoted to investigate lactate metabolism under various conditions in vivo [3][4][5][6] , underlying cellular processes are still not well described, and studies are mainly limited to cell lines 7,8 . Expanding knowledge on the cellular metabolism of lactate, as well as its impacts on fuel utilization of other energy substrates in skeletal muscle cells, is of importance, as these processes will affect whole-body energy homeostasis.
Depending on the level of contractile activity, energy consumption in skeletal muscles may vary considerably, with an estimated 1000-fold difference between rest and maximal contraction 9 . Energy required for cellular functions, including contraction, is provided by ATP hydrolysis to ADP and P i . There are three main mechanisms to generate ATP in skeletal muscle; creatine kinase activity, glycolysis and mitochondrial oxidative phosphorylation. The different muscle fiber types differ in their metabolic preferences. Slow muscles (type I fibers) rely mainly on oxidative phosphorylation, whereas fast muscles (type II fibers) are primarily glycolytic. Plasma lactate concentrations increase rapidly during exercise, and the traditional idea has been that lactate is a waste product that is transported to the liver, kidneys or other organs for clearance 1 . However, it is now generally accepted that lactate is also taken up by the muscles and oxidized 10 . In fact, the contribution of muscles to total body lactate clearance is considerable during exercise [10][11][12] . Most lactate (75-80%) is disposed of immediately within the tissue or released for reuptake by working muscle 13 . Even in resting muscles, when partial pressure of oxygen is more than sufficient for oxidative phosphorylation to occur, concentrations of lactate and pyruvate are 1.0 mM and 0.1 mM, respectively 14 . After vigorous exercise, but still at oxygen pressure sufficient for maximal mitochondrial

Results
Lactate as an energy substrate for myotubes. Extracellularly added lactate must be transported into muscle cells in order to function as an energy substrate. Expression levels of the two main lactic acid transporters in skeletal muscles, MCT1 (SLC16A1) and MCT4 (SLC16A4), were studied in fully differentiated myotubes grown in standard differentiation medium (5.5 mM glucose) at both mRNA (Fig. 1a) and protein level (Fig. 1b,c). Two other transporters, MCT2 and MCT3, were studied at protein level (Fig. 1b,c). mRNA expression level of MCT4 (SLC16A4) was reduced in cells treated with 4 mM lactic acid (Fig. 1a), while protein levels of other MCTs were unaffected after 24 h of lactic acid pretreatment, except for MCT2, which was reduced in cells treated with 10 mM lactic acid (Fig. 1b,c).
In addition to be taken up and oxidized, lactic acid was found to be a substrate for glycogen synthesis in human myotubes (Fig. 3), and insulin increased incorporation of lactic acid into glycogen in four out of six cell donors (Fig. 3).
Lactic acid was also incorporated into cellular lipids (Fig. 4), with a total lipid content of 57.9 ± 29.5 nmol/mg protein. Incorporation into phospholipids (PL) constituted approximately 70% of total lipid content (Fig. 4). Diacylglycerol (DAG) and triacylglycerol (TAG) constituted approximately 10% and 20% of the total lipid content, respectively (Fig. 4), whereas both free fatty acids (FFA) and cholesteryl ester (CE) constituted approximately 2% of the total lipid content (Fig. 4). After hydrolysis of TAG and DAG, 90% of the incorporated lactic acid was recovered in the aqueous (glycerol) phase, whereas 10% was recovered in the organic (fatty acids) phase (data not shown).
Effects of lactate on the metabolism of glucose and oleic acid. Acute effects of lactate exposure.
Acute addition of lactic acid (4 h) did not affect glucose uptake (Fig. 5a), whereas glucose oxidation was decreased when lactic acid was added at the highest concentration (10 mM, Fig. 5b). However, acutely added lactic acid had no effect on fractional glucose oxidation (Fig. 5c). Lactic acid increased oleic acid uptake (Fig. 5d), whereas oleic acid oxidation was markedly reduced (Fig. 5e), causing reduced fractional oleic acid oxidation at all concentrations (Fig. 5f).
Chronic effects of lactate exposure. Chronic exposure to lactic acid (5 mM for 24 h), compared to chronic exposure to the same concentration of glucose, revealed a tendency towards enhanced glucose uptake (Fig. 6a), whereas oleic acid metabolism was unaffected (Fig. 6b). Furthermore, protein expressions of the glucose Figure 1. mRNA and protein expression of lactate transporters. Satellite cells were cultured and differentiated to myotubes. They were kept in normal differentiation medium (basal, 5.5 mM glucose) or exposed to lactic acid (4 mM or 10 mM replacing glucose) for 24 h before total cell RNA and protein content were isolated. (a) mRNA expressions of SLC16A1 (MCT1) and SLC16A4 (MCT4). All values were corrected for the housekeeping gene acidic ribosomal phosphoprotein P0 (RPLP0), and presented as means ± SEM relative to basal (n = 5). (b,c) Protein expressions of MCT1-4. (b) One representative immunoblot. (c) Quantified immunoblots relative to basal. All values were corrected for the housekeeping control α-tubulin, and data are presented as means ± SEM relative to basal (n = 5). Dividing lines delineate blots from different gels. The samples derive from the same experiment, and all blots were processed in parallel. Full-length blots are presented in Supplementary Fig. 1 transporter GLUT4, as well as the fiber type markers MHCI (type I fibers) and MHCIIa (type II fibers) were unaffected by lactic acid (4 mM or 10 mM) exposure for 24 h (Fig. 6c,d).
Effects of prolonged lactate exposure. When increasing the exposure time by replacing glucose with lactic acid during the whole proliferation and differentiation period; both glucose uptake (Fig. 7a) and oxidation (Fig. 7b) were increased. Oleic acid uptake (Fig. 7c) was not affected, but oxidation was enhanced (Fig. 7d). The effect of lactic acid was comparable to removing glucose completely (0 mM glucose, Fig. 7). Replacing glucose with lactic acid only during the differentiation period (7 days) did not affect oleic acid metabolism (data not shown).
To further investigate underlying mechanisms in conditions where glucose was replaced by lactic acid for the entire culturing period, we studied mRNA and protein expressions of several factors involved in metabolic processes in skeletal muscles (Fig. 8). Compared to the cells grown in glucose-containing media, mRNA expressions of the transporters of glucose and fatty acids, solute carrier family 2 member 4, GLUT4 and CD36, respectively,  were decreased in cells cultured in lactate-containing media (Fig. 8a). On the other hand, carnitine palmitoyl transferase 1B (CPT1B) and cytochrome c1 (CYC1), involved in the transport of fatty acids across the mitochondrial membrane and mitochondrial function, respectively, were increased in lactate-cultured cells (Fig. 8a). mRNA expression levels of ANGPTL4, a potent inhibitor of lipoprotein lipase activity induced in the fasting state by peroxisome proliferator activated receptors (PPARs) 37 , CPT1A and pyruvate dehydrogenase kinase 4 (PDK4), an inhibitor of the pyruvate dehydrogenase complex, were not changed in cells cultured in the presence of lactate (Fig. 8a). Protein expression of the electron transport chain's complex V (ATP synthase subunit α) was also increased in the presence of lactic acid (Fig. 8b).
Despite the observed changes in the metabolism of both glucose and fatty acids, the metabolic flexibility of myotubes grown and differentiated in lactic acid-containing media did not differ from cells grown in standard glucose-containing medium, assessed as glucose suppression of oleic acid oxidation (suppressibility) and the ability to increase oleic acid oxidation with increasing oleic acid concentration (adaptability) (data not shown).

Discussion
The aim of the present study was to explore the ability of cultured human myotubes to utilize lactate as a fuel source, but also to investigate effects of lactate on the metabolism of the two other main energy substrates, glucose and oleic acid. Lactate is no longer considered as only a waste product of glycolysis due to hypoxia, but also an important factor in cellular and whole body metabolism. Glucose and fatty acids are the two most important fuel sources of skeletal muscle. Moreover, skeletal muscle is the largest producer of lactate. Thus, shedding light on the impact of lactate on cellular metabolism of glucose and fatty acids in skeletal muscle cells is of particular interest for understanding whole-body energy balance, and has, to our knowledge, not been thoroughly explored. Our main finding was that isolated human skeletal muscle cells have the ability to metabolize lactate in terms of uptake, oxidation and storage not only in form of glycogen, but also as intracellular lipids. Furthermore, prolonged exposure to lactic acid increased glucose metabolism and oleic acid oxidation. These findings were accompanied by increased mRNA expression levels of CPT1B and CYC1, as well as higher protein expression of complex V (ATP synthase subunit α) of the mitochondrial respiratory chain, suggesting an overall improvement in mitochondrial function of myotubes exposed to lactic acid for an extended period.
The ability of the cells to take up extracellularly added lactate was first verified by showing expression of the monocarboxylate transporters responsible for lactate transport in skeletal muscles, MCT1 and MCT4, both at mRNA and protein level, and MCT2 and MCT3 at protein level. It has recently been shown in C2C12 cells that 16 mM lactate induced an increase in mRNA expression levels of both SLC16A1 (MCT1) and SLC16A4 (MCT4), but with different times of peak induction, suggesting different dynamics in the expression of the two transporters upon lactate stimulation 38 . We observed lower mRNA expression of SLC16A4 in cells pretreated with 4 mM lactic acid for 24 h, but not with 10 mM lactic acid. Lactate concentrations may reach 10-20 mM in the circulation 39,40 , or even 40 mM in the skeletal muscle after intense anaerobic exercise in humans, which has been shown to increase expression of both lactate transporters 41 . In another in vivo study, a progressive decrease in mRNA expression levels of both MCT1 and MCT4 over the period of 9-72 h after a single bout of exercise was reported 42 . Interestingly, in the same study, protein content of MCT1 was higher after 24-72 h compared to 9 h post-exercise, while MCT4 remained unchanged 42   biopsies post-exercise or cell-harvesting to interpret adaptive responses. Our finding of decreased mRNA level of MCT4 at 4 mM of lactic acid is supportive to in vivo results, and in the present study, both protein and RNA isolation was performed immediately after ended lactic acid treatment. The fact that the observed decrease in mRNA expression of SLC16A4 at 4 mM of lactic acid was not seen when concentration was increased to 10 mM is somewhat unexpected. However, it has been shown that MCT4 is upregulated in vitro by hypoxia-inducible factor-1α pathway, and hypoxia during exercise has been suggested to induce transient mRNA bursts of MCT4, followed by subsequent reduction in mRNA contents 43 . Thus, discrepancies between mRNA and protein levels, as well as inconsistencies regarding effects of exercise in vivo studies have been reported earlier, suggesting complex dynamics and post-transcriptional modifications that are yet to be described. The concentrations used in this study were in the range 2-10 mM, and at these conditions, no differences in mRNA expression of SLC16A1 or protein expressions of MCT1 or MCT4 were observed. As mentioned in the introduction, MCT1 has predominantly been found in oxidative fibers, whereas MCT4 is more abundant in glycolytic fibers 19,44 . Cultured human myotubes are generally limited with regards to their reflection of the fiber type of the muscle they are isolated from, and tend to mainly express fiber type markers of fast, glycolytic fibers 45 . It has also been shown that peroxisome proliferator activated receptor gamma coactivator −1α (PGC-1α) is a strong inducer of both MCT1 (but not MCT2 and MCT4) expression and uptake of lactic acid in muscles 46 . Expression of PGC-1α, which is a contraction-induced regulator of oxidative mitochondrial function, is reduced in cultured human myotubes 47 . Low expression of PGC-1α could partially explain lack of increase in protein expression of MCT1 in our study. We further found that both MCT2 and MCT3 were expressed in cultured human myotubes. Interestingly, protein level of MCT2 was reduced in myotubes exposed to the highest concentration of lactic acid (10 mM). Whether MCT2 adapts to chronically increased muscle activity is still not defined, although it was reported in one study that it is not regulated by PGC-1α, nor does it change in response to contractions 46 . MCT2 is expressed in skeletal muscles and it is primarily involved in import of lactic acid, with a higher affinity for both lactic acid and pyruvate than MCT1 46 . This makes it a suitable transporter in tissues taking up large amounts of lactic acid as a fuel. In spite of that, expression of MCT2 does not appear to correlate to oxidative capacity of different muscles, nor is it more abundant in oxidative muscles, in contrast to MCT1 46 . Mechanisms of regulation of MCT2 are also not clarified. MCT3, which is even less described, was also expressed in myotubes in the present study, but protein expression of this transport was unaffected by different concentrations of lactic acid. Higher abundance of MCT3 in fast-twitch glycolytic and fast-twitch oxidative than in slow-twitch muscle fibers has been reported, and it appears to be involved in the efflux of lactic acid 25 . It is important to keep in mind that in vivo, most muscles contain a mixture of both primarily glycolytic and primarily oxidative muscle fibers, and the proportion of each fiber type will depend on many factors from genetics to whether that particular muscle uses high-intensity exercise (glycolytic) or endurance exercise (oxidative). Thus, both expression and regulation of the MCTs is probably tightly associated with the particular muscle and its metabolic activity.
The fact that all four transporters were expressed in the cells suggests that cultured myotubes should be able to transport lactate across membranes. This was confirmed by flux studies of labelled lactate, which was shown to be taken up by the cells, and approximately 45% was further oxidized to CO 2. This is in accordance with the in vivo findings, where it has been shown that muscle fibers can take up lactate for subsequent metabolic use of intramuscular lactate 48,49 . Oxidative muscle fibers predominantly oxidize lactate, whereas glycolytic fibers primarily convert lactate to glycogen [50][51][52][53] . A probable pathway of intracellular lactate disposal involves conversion to pyruvate followed by entry into the Krebs cycle 54 . We further show that, in addition to being oxidized, lactate can be incorporated and stored as glycogen as well. This is in accordance with in vivo studies; glyconeogenesis from lactate has been observed in human muscles 55 . Studies on rat muscle showed both glyceroneogenesis and glyconeogenesis from lactate, and reverse flux of pyruvate to phosphoenolpyruvate was found to be the common route 56 . Indeed, during exercise, lactate is by far the most important gluconeogenic precursor in humans, as it is during fasting 57 .
Further, we showed that lactate can be incorporated in complex intracellular lipids. Chen et al. showed that lactate was metabolized to lipids in cultured HeLa cells and H460 human lung cancer cells 58 , and Jin et al. showed in studies on rat muscle that lactic acid could be converted to glycerol 56 . It has also been reported that lactate, at physiological concentrations, can induce triglyceride accumulation in adipocytes of both humans, mice and rats 59 , and recently, a dose-dependent lactate-induced induction of triglyceride content was demonstrated in C2C12 cells at concentrations of 16 mM and 20 mM 38 . As far as we know, incorporation of lactate to lipids has never before been observed in human skeletal muscle cells. While the majority of lactic acid was incorporated into phospholipids in the present study, partition to TAG was also substantial. Of the lactic acid incorporated in DAG and TAG, 90% was recovered in the aqueous (glycerol) phase, whereas 10% was recovered in the organic (fatty acid) phase. We further found that pyruvate decreased both uptake and oxidation of lactic acid, as would be expected of a pyruvate metabolite. Glucose was also found to decrease both uptake and oxidation of lactate, which is in line with reports of human myotubes being highly glycolytic when grown in presence of glucose 60 , in addition to having low mitochondrial oxidative capacity resembling fast (glycolytic) muscle fibers 61 . Cultured human myotubes' reliance on glucose as favorable fuel is also demonstrated by our observations that higher concentrations of acutely added lactate were necessary to suppress glucose oxidation than oleic acid oxidation (10 mM vs. 4 mM, respectively). Palmitic acid also caused a decrease in lactate oxidation. The monocarboxylic acid transport inhibitor CHC caused, as expected, a decrease in lactate uptake. Insulin had no effect on lactate metabolism and cytochalasin B did not inhibit lactate uptake, implying an uptake mechanism independent of translocation of transporters.
Acutely added lactate (4 h) suppressed both glucose and oleic acid metabolism, which is in agreement with a previous study where increased plasma lactate was associated with a decline in plasma FFAs, an anti-lipolytic effect on adipose tissue, and an inhibitory effect on muscle fat oxidation 54 . Chronic treatment (24 h) with 5 mM lactic acid, on the other hand, had no effect on either glucose or oleic acid metabolism; neither did it affect expression of the glucose transporter GLUT4. In these cells, glucose in the culturing media was replaced with the same concentration of lactate (5 mM) for a period of 24 h prior to harvesting of the cells, but not during the experiment. Therefore, possible explanations for the lack of effects may be too short pretreatment and absence of lactate during the 4 h course of experiment.
To increase exposure time to lactate, cells were exposed to either lactate or glucose for the entire culturing period. At these conditions, both glucose uptake and oxidation were increased, whereas oleic acid oxidation was increased without a corresponding effect on uptake. Thus, by replacing glucose with lactate for the entire culturing period, we induced an increase in the oxidative metabolism of the cells, which was mainly reflected as an increase in glucose metabolism. Based on these results, one may suggest that lactate forces the cells towards a state of energy deprivation and thereby increases the oxidative capacity of the cells, as metabolism levels were fairly similar as after culturing in media containing no glucose. and cytochrome c1 (CYC1). All values were corrected for the housekeeping control RPLP0, and presented as means ± SEM relative to basal (n = 5). (b,c) Expressions of proteins involved in mitochondrial phosphorylation, OXPHOS (complex I subunit NDUFB8, complex II subunit, complex III subunit core 2, complex IV subunit II and complex V (ATP synthase subunit α)). (b) One representative immunoblot. (c) Quantified immunoblots normalized to glucose. All values were corrected for the housekeeping control α-tubulin, normalized to glucose, and data are presented as means ± SEM relative to basal (n = 5). Dividing lines delineate blots from different gels. The samples derive from the same experiment, and all blots were processed in parallell. Full-length blots are presented in Supplementary Fig. 8 To shed light on the underlying mechanisms of the observed effects, we studied protein and mRNA expression levels of several metabolically important factors. Compared to the cells grown in glucose-containing media, mRNA expressions of the transporters of glucose and fatty acids, GLUT4 and CD36, respectively, were decreased in cells cultured in presence of lactic acid. On the other hand, expressions of two important markers of mitochondrial function and fatty acid oxidation, CYC1 and CPT1B, respectively, were both increased after prolonged exposure to lactate, as was protein expression of complex V (ATP synthase subunit α) of the mitochondrial respiratory chain. These data generally support functional results and suggest an increased transport of the fatty acids taken up across the inner mitochondrial membrane, an improved mitochondrial function and establishment of a more oxidative cell model in general. However, decreased level of GLUT4 was in contrast to the observed increase in glucose uptake, but this observation can probably be explained by generally low GLUT4 expression in skeletal muscle cells in vitro 62 , and the fact that basal glucose uptake can mainly be mediated by other glucose transporters (GLUT1 and 3) in primary human myotubes 63,64 . Discrepancies between GLUT4 levels and glucose uptake have also been reported previously 47 .
We have previously observed a remodeling of oxidative energy metabolism by replacing glucose with galactose as carbohydrate source during growth and differentiation 65 . In general, myotubes became more oxidative and seemed to utilize glucose better than oleic acid, which we suggested implied an improved metabolic switching 65 . In the present study, however, none of the parameters of metabolic switching were affected in cells that had been cultured under conditions where glucose was replaced by lactate. Thus, although replacing glucose with lactate had a positive effect on both oleic acid and glucose oxidation, the metabolic flexibility of the cells was unaffected.
In conclusion, cultured human skeletal muscle cells are able to take up, oxidize and store lactate in complex lipids and glycogen. Further, we have found that lactate exposure affects the metabolism of glucose and oleic acid in cultured human myotubes, and this impact appears to be related to the duration of exposure to lactate. These findings suggest an important role of lactate in energy metabolism of human skeletal muscle cells.  66 , MCT1 (H-1) (sc-365501) 67 and MCT4 (G-7) 68 (sc-376465) were from Santa Cruz Biotechnology (Santa Cruz, CA, US). The antibody against MHCI (MAB1628) 69 was from Millipore (Temecula, CA, US). The human total OXPHOS antibody (ab110411) 65 and antibodies against MCT2 (ab129290) 70 and SLC16A8 (ab60333) 70  Ethics statement. The biopsies were obtained with informed written consent and approval by the Regional Committee for Medical and Health Research Ethics South East, Oslo, Norway (reference number: S-04133). The study adhered to the Declaration of Helsinki.

Materials. Dulbecco's Modified
Human skeletal muscle cell cultures. Satellite cells were isolated as previously described 73  humidified 5% CO 2 atmosphere at 37 °C, and medium was changed every 2-3 days. Growth and differentiation of cells in media containing lactic acid was done by adding DMEM-Glutamax with no glucose, 5 mM L-glutamine, 5 mM lactic acid, 2% FCS, 2% Ultroser G, penicillin/streptomycin, gentamycin, and amphotericin B the day after seeding. At about 70-80% confluence, medium was changed to DMEM with no glucose, 5 mM glutamine, 5 mM lactic acid, 2% FCS, 25 pM insulin, penicillin/streptomycin, gentamycin, and amphotericin B. For comparison, cells were also grown in regular culturing and differentiation media (DMEM-Glutamax, 5.5 mM glucose) without lactic acid.
Lactate treatments. Different concentrations of lactate were used for the purpose of specific experiments in the present study. Initially, the ability of the myotubes to take up and oxidize lactate was assessed by using L-[ 14 C(U)]lactic acid (1 µCi/ml, 100 μM). To verify protein and mRNA expression of the two lactate transporters, MCT1 and MCT4 (representative genes SLC16A1 and SLC16A4, respectively), myotubes were treated with 4 mM or 10 mM lactate for 24 h. Ability to store lactate as glycogen and intracellular lipids, was determined using L-[ 14 C(U)]lactic acid (1 µCi/ml, 20 mM) for 24 h. Concentration of 20 mM was decided based on previous studies investigating incorporation of lactate into triglycerides and glycogen 8,51 .
To investigate impacts of lactate exposure on metabolism of glucose and oleic acid, we differed between acute, chronic and prolonged lactate treatment. In the acute treatment (4 h), myotubes were exposed to increasing concentrations of lactate (2 mM, 4 mM, 6 mM and 20 mM), covering the range of concentrations from physiological levels to those seen after exercise in vivo 74,75 . In chronic lactate treatment, glucose in the culturing medium was replaced by the same amount of lactate (5 mM) for 24 h. Finally, prolonged exposure was determined as a condition where glucose in the culturing media was completely replaced by the same concentration of lactate (5 mM) for the entire proliferation and differentiation period. (1 µCi/ml), was given during 4 h CO 2 trapping as described previously 76 . Substrates were added in DPBS with 10 mM HEPES, 10 µM BSA and 1 mM L-carnitine (L-carnitine was only added with oleic acid). Oleic acid was bound to BSA at a ratio of 2.5:1 (this was adjusted for in the substrate medium). A 96-well UniFilter ® microplate, soaked with NaOH (1 M), was mounted on top of the CellBIND ® plate and produced CO 2 was trapped during 4 h incubation at 37 °C. CO 2 production was measured in the presence or absence of various compounds: insulin (100 nM), glucose (5 mM), oleic acid (100 µM), palmitic acid (100 µM), pyruvate (5 mM), cytochalasin B (10 µM), or CHC (1 µg/ml). CO 2 production and cell-associated (CA) radioactivity were assessed using a PerkinElmer 2450 MicroBeta 2 scintillation counter. Protein content in each well was determined with the Bio-Rad protein assay using a VICTOR ™ X4 Multilabel Plate Reader as described previously 77 . The sum of 14 CO 2 and CA was considered as total substrate uptake. Fractional complete oxidation was calculated as CO 2 /(CO 2 + CA).

Metabolic flexibility parameters.
Suppressibility is the ability of the cells to decrease oleic acid oxidation by acutely added glucose, and is calculated as [1 − (oxidation of 100 µM OA at 5 mM glucose/oxidation of 100 µM OA at no glucose added) × 100%] 33 . Adaptability is the ability to increase oleic acid oxidation with increasing oleic acid concentration, and is calculated as [oxidation of 100 µM OA/oxidation of 5 µM OA] 33  Myotubes were then washed twice with PBS and harvested with 200 μl of 0.1% SDS. Cellular lipids were extracted, as previously described 78 , from homogenized cell fractions, separated by thin layer chromatography (TLC) and quantified by liquid scintillation (Tri-Carb 1900, PerkinElmer). A non-polar solvent mixture of hexane:diethyl ether:acetic acid (65:35:1) was used to separate the lipids. The amount of lipids was related to total cell protein concentrations. Saponification of the total lipids was performed as described in 79 with minor modifications. The cellular total lipid extract from DAG and TAG, respectively, were dried and dissolved in 0.75 ml of 30% KOH and heated at 70 °C for 10 min. An equal volume of 95% ethanol was then added and the mixture was heated at 70 °C for 2 h. After cooling, the aqueous fraction was acidified with 3 M HCl drop-wise and extracted three times with 2 ml light petroleum. The organic fraction was evaporated, redissolved in 1 ml of light petroleum, transferred to scintillation vials and quantified as the acyl fraction using the Tri-Carb scintillation counter. The remaining aqueous fraction contained the glyceryl fraction; 0.5 ml aliquot of this sample was taken and measured for radioactivity using the TriCarb scintillation counter.
Glycogen synthesis. Myotubes were exposed to differentiation media supplemented with L-[ 14 C(U)]lactic acid (1 µCi/ml, 20 mM) for 24 h prior to assay start. At assay start, myotubes were exposed to serum-free DMEM supplemented with L-[ 14 C(U)]lactic acid (1 µCi/ml, 20 mM), in presence or absence of 100 nM insulin (Actrapid ® Penfill 100 IE/ml), for 3 h to measure glycogen synthesis. The cells were washed twice with PBS and harvested in 1 M KOH. Protein content was determined by use of the Pierce BCA Protein Assay Kit, before 20 mg/ml glycogen and more KOH (final concentration 4 M) were added to the samples. Then L-[ 14 C(U)]lactic acid incorporated into glycogen was measured as previously described 33 . Immunoblotting. Myotubes were cultured in 6-well plates, and two parallell wells of each donor were treated with either 4 mM or 10 mM lactic acid. Samples for immunoblotting were harvested in Laemmli buffer, and aliquots of 15 μg cell protein were electrophoretically separated by SDS-PAGE (Bio-Rad 4-20% Mini Protean ® TGX ™ precast gels with Tris/glycine buffer) and transferred to nitrocellulose membranes. The membranes were (version 4.0) software. All samples were derived at the same time and processed in parallel. Expression levels were normalized to one sample used as loading control and further normalized to the endogen control α-tubulin (1:1000).

RNA isolation and analysis of gene expression by qPCR.
Human skeletal muscle cells were washed, trypsinized and pelleted before total RNA was isolated using QIAGEN RNeasy Plus Mini Kit according to the supplier's protocol. Total RNA was reversely transcribed (25 °C for 10 min, 37 °C for 80 min, 85 °C for 5 min) with a High-Capacity cDNA Reverse Transcription Kit and TaqMan Reverse Transcription Reagents using a 2720 Thermal Cycler. Primers were designed using Primer Express ® and real-time qPCR was performed using a StepOnePlus Real-Time PCR system. Each target were quantified in triplicates and carried out in a 25 µl reaction volume in accordance with the supplier's protocol. All assays were run for 44 cycles (95 °C for 15 s followed by 60 °C for 60 s). The transcription levels were normalized to the housekeeping gene acidic ribosomal phosphoprotein P0 (RPLP0, acc.no. M17885). The housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH, acc.no. NM_002046) was also analyzed; there were no differences between normalizing for RPLP0 or GAPDH. The following forward and reverse primers were used at concentration of 30 µM: ANGPTL4 (acc. no. NM_139314), CD36 (acc.no. L06850), CPT1A (acc.no. L39211), CPT1B (acc.no. D1852C12) and CYC1 (acc. no. NM_001916), GAPDH, GLUT4 (acc.no. M20747), PDK4 (acc.no. BC040239), RPLP0, SLC16A1 (acc.no. NM_003051, representing MCT1), and SLC16A4 (acc.no. NM_004696, representing MCT4).
Statistics. Data are presented as means ± SEM in nmol/mg protein or as percent of control. The value n represents the number of different donors, each with at least duplicate observations. Statistical comparison between different treatments was performed by Student's t-test using GraphPad Prism 6 for Windows. The parameter of interest was entered as the dependent variable and pretreatment (lactic acid or no glucose) and acute treatments were entered as fixed variables. Differences were considered statistically significant at p-values < 0.05.