Introduction
Skeletal muscle tissue plays a pivotal role in the regulation of whole body metabolism, functional work capacity, and exercise performance (Zurlo et al., 1994; Egan and Zierath, 2013). While no medicine has proven to be more efficient than exercise to improve muscle function (Booth et al., 2000), exercise-nutrition interactions significantly impact on energy substrate replenishment and on muscle remodeling and repair post-exercise (Beelen et al., 2010; Hawley et al., 2011). During recovery from exercise, amino acid intake is required to stimulate muscle protein synthesis (Biolo et al., 1997; Churchward-Venne et al., 2012), whilst high-rate carbohydrate intake promotes resynthesis of glycogen stores (Costill et al., 1981; Burke et al., 2016). Furthermore, improving the (nutritional) conditions under which exercise promotes muscle protein synthesis is not only important for athletes, but also in clinical conditions requiring enhanced anabolic response in patients who are unable to perform high-intensity exercise.
Skeletal muscle protein synthesis following exercise is mainly regulated by the mechanistic target of rapamycin complex 1 (mTORC1) (Bodine et al., 2001; Schiaffino and Mammucari, 2011). mTORC1 regulates protein translation initiation and elongation via phosphorylation of the ribosomal protein S6 kinase 1 (S6K1) and the eukaryotic initiation factor 4E-binding protein 1 (4E-BP1) (Burnett et al., 1998). Muscle contractions per se activate mTORC1 in a stress and/or mechanic-related manner, but also the intake of proteins activates mTORC1 and protein synthesis (Morton et al., 2015). On the other hand, as protein synthesis is a high ATP-consuming process, mTORC1 activity is inhibited during periods of energetic stress by the AMP-activated protein kinase (AMPK; Bolster et al., 2002; Dreyer et al., 2006). AMPK is the principal energy sensor in muscle cells and is activated by decreased energy levels during and after contractions (Hardie, 2011). AMPK promotes energy substrate catabolism to generate ATP, and at the same time inhibits anabolic processes requiring ATP, in order to preserve energy status (Gwinn et al., 2008). During and after strenuous exercise, activation of AMPK is responsible for the delayed activation of muscle protein synthesis (Dreyer et al., 2006; Thomson et al., 2008). Rapid restoration of muscle energy balance to de-activate AMPK is important in that it enhances post-exercise mTORC1 activation, protein synthesis and eventually training-induced muscle remodeling.
Muscle glycogen is the primary fuel for muscle contractions during high-intensity endurance exercise. Glycogen stores are small relative to the total energy need in prolonged exercise, which makes initial muscle glycogen content an important determinant of performance (Hawley et al., 1997). Hence, maximizing muscle glycogen resynthesis is crucial for maintenance of endurance exercise performance whenever time for recovery between events is short relative to the time needed to full glycogen repletion (Burke et al., 2016).
Acetoacetate (AcAc) and beta-hydroxybutyrate (βHB) are the main ketone bodies (KB) produced in the liver in response to low blood glucose and insulin levels, and play a critical role in survival during episodes of fasting. Three weeks of starvation result in blood ketone levels of up to ~6 mM (Cahill, 2006). Alternatively, a strict ketogenic diet containing low carbohydrate, low protein, but high fat (Cahill and Veech, 2003), can also elevate circulating KB levels to ~1–2 mM within 2–4 days (Pinckaers et al., 2016). Under those conditions, KB serve as circulating energy substrates for metabolic tissues such as the brain (Kashiwaya et al., 2010), cardiac (Nakatani et al., 2015), and skeletal muscle (Cahill, 2006). Mechanistically, KB fuel the tricarboxylic acid (TCA) cycle after being converted to acetyl-coenzyme A (Shi and Tu, 2015; Murray et al., 2016), and prevent the depletion of carbohydrate stores as well as the breakdown of muscle contractile proteins for gluconeogenesis and TCA cycle anaplerosis. While KB in this way could potentially affect performance, the dietary conditions needed to stimulate endogenous KB synthesis are not compatible with optimal performance in most sports (Cox and Clarke, 2014). Recently, the ketone monoester (R)-3-hydroxybutyl (R)-3-hydroxybutyrate (hereafter referred to as KE) has emerged as an alternative approach to raise blood KB levels in the absence of any other dietary modification. Ingestion of KE increases blood ketone appearance within minutes of consumption (Clarke et al., 2012b), leading to circulating concentrations similar to those observed in prolonged fasting, even in the presence of elevated blood insulin levels (Clarke et al., 2012b; Cox et al., 2016). KE enhanced endurance exercise performance in athletes, at least partly because KB act as an alternative substrate for oxidative phosphorylation (Cox et al., 2016). Interestingly, by reducing the need for glucose-derived ATP production, KE also resulted in glycogen sparing during exercise (Cox et al., 2016).
Against the above background, we decided to investigate whether nutritional ketosis, induced via KE ingestion, could enhance markers of muscle protein synthesis and/or glycogen resynthesis during recovery from strenuous exercise. While ingestion of KE did not increase muscle glycogen resynthesis, it increased activation of muscle protein synthesis markers. Using an in vitro strategy, we confirmed that KB potentiated leucine-mediated mTORC1 activation and that this resulted into higher rate of leucine-mediated protein synthesis. Altogether, our data show that KE enhances mTORC1 activation during recovery from exercise but does not affect muscle glycogen resynthesis in human muscle.
Discussion
In this study, we investigated the potential of nutritional ketosis to stimulate markers of post-exercise muscle protein and glycogen synthesis, since both may enhance long-term training adaptations and facilitate muscle recovery from intense exercise (Hawley et al., 2011). We found that oral KE ingestion increased mTORC1 activation in human skeletal muscle following strenuous exercise, but left muscle glycogen resynthesis unaffected. To explore whether the increase in mTORC1 activation leads to increased protein synthesis, we subsequently performed a set of in vitro experiments using C2C12 myotubes. We confirmed that ketone bodies stimulated mTORC1 activity and showed that this leads to enhanced leucine-mediated muscle protein synthesis. Taken together, our results indicate that KE enhances the anabolic response to both exercise and protein ingestion but does not affect glycogen resynthesis.
We aimed to elevate plasma ketone concentrations using intermittent (60 min) drinks of a novel oral ketone ester (Clarke et al., 2012b) during the initial 5 h of recovery following a glycogen-depleting maximal exercise bout. In conjunction with KE, we also administered a mixture containing whey protein and carbohydrates in a quantity that is considered to elicit maximal rates of post-exercise muscle protein (Morton et al., 2015) and glycogen synthesis (Burke et al., 2016), because we wanted to explore whether KE could add to the recommended nutritional procedures for stimulation of muscle recovery. KE increased blood βHB concentration to ~3–5 mM within 30 min after the first drink to remain elevated to the end of the recovery period in all subjects. As KE raises blood βHB:AcAc levels in a 4:1 ratio (Clarke et al., 2012b) total plasma KB concentrations above 6 mM are conceivable, levels similar to those reported after several weeks of starvation (Cahill, 2006).
We did not detect any rise in post-exercise ketone levels in the PL condition. Post-exercise ketosis is inversely correlated to liver glycogen content (Adams and Koeslag, 1988), which in short one-leg postprandial exercise protocols is not depleted (Gonzalez et al., 2016). In addition, post-exercise nutritional intake, causing hyperinsulinemia, also reduces endogenous ketosis (Koeslag, 1982; Impey et al., 2016). Therefore, in PL post-exercise ketosis is unlikely to have significantly contributed to muscle recovery.
Muscle protein synthesis following exercise is mainly controlled by mTORC1 (Bodine et al., 2001; Schiaffino and Mammucari, 2011). Assessment of the phosphorylation status of the mTORC1 substrates S6K1 and 4EBP1 is a reliable readout of mTORC1 activity and is closely linked to muscle hypertrophy both in humans (Terzis et al., 2008) and mice (Baar and Esser, 1999). Moreover, S6K1 activation is required to increase muscle force (Marabita et al., 2016). We found the ingestion of KE to increase mTORC1 activity during recovery from exercise, as indicated by higher phosphorylation status of S6K1 and 4EBP1 (Figure 4). While evaluation of muscle protein synthesis via key proteins involved in protein translation has been shown to be a reliable procedure (Baar and Esser, 1999; Bodine et al., 2001; Terzis et al., 2008; Drummond et al., 2009), a dissociation between S6K1/4EBP-1 phosphorylation and muscle protein synthesis might still be possible (Greenhaff et al., 2008). Since our study protocol did not allow us to assess muscle protein synthesis rates in human subjects in vivo, we decided to set up a series of in vitro experiments using C2C12 myotubes. First, we confirmed that KB potentiate mTORC1 signaling following leucine stimulation (Figure 7). Subsequently, we found that the increase in mTORC1 activity was reflected by increased protein synthesis in muscle cells in vitro (Figure 8). While our in vitro data show the ability of KB to potentiate leucine-mediated protein synthesis in myotubes, further human studies using stable isotope tracers need to confirm whether or not KE can also promote muscle protein synthesis in humans.
The ability for KE to promote mTORC1 activity occurred in the absence of increased blood insulin levels and downstream insulin signaling pathway. Moreover, we could reproduce these effects using strictly controlled in vitro settings. This indicates that the effect of KE/KB is independent of growth factor signaling as well as potential other systemic effects induced by ketosis (Egan and D'Agostino, 2016; Holdsworth et al., 2016).
While we for the first time show that KE stimulates mTORC1 signaling post-exercise against the background of high-rate protein plus carbohydrate intake in humans, it is well established that KB modulate muscle protein metabolism during starvation. When glucose availability is low, KB replace glucose as the primary brain fuel, and therefore suppresses the need for muscle proteolysis to provide gluconeogenic precursors (Sherwin et al., 1975; Owen, 2005). Furthermore, hyperketonaemia induced by intravenous infusions of a ketone salt after an overnight fast, decreased leucine oxidation and increased protein synthesis in humans (Nair et al., 1988). The direct effect of KB on protein synthesis rates in muscle has only been explored in a limited number of ex vivo studies. One study found physiological concentrations of KB without amino acids in the incubation medium not to stimulate protein synthesis in myocytes (Tischler et al., 1982).
Another ex vivo study tested the effect of either 4 mM βHB or 4 mM AcAc in chicken muscles and reported a decrease in protein synthesis. However, raising glutamine availability restored protein synthesis (Wu and Thompson, 1990). By analogy, our results show a synergistic effect of physiological concentrations of ketone bodies and leucine on S6K1 phosphorylation (Figure 7) and a direct effect of ketone bodies on leucine-mediated protein synthesis (Figure 8). Taken together these results indicate the potential of ketone bodies to stimulate muscle protein anabolism.
Although we are unable to identify the precise mechanism responsible for enhanced muscle protein synthesis with KE, it is known that skeletal muscle protein synthesis is reduced when energy homeostasis is impaired (Hardie, 2011). Following high-intensity exercise, activation of AMPK is responsible for the delayed activation of muscle protein synthesis (Dreyer et al., 2006; Thomson et al., 2008). Rapid restoration of muscle energy balance to de-activate AMPK is therefore important to initiate post-exercise mTORC1 activation and protein synthesis. KE ingestion markedly promoted the decrease in AMPK activation early during recovery, potentially via serving as an additional energy substrate for oxidative ATP production (Cox et al., 2016; Murray et al., 2016) which might allow faster recovery of energy homeostasis and facilitate mTORC1 activation.
Nutritional ketosis was recently shown to alter substrate competition during endurance exercise and to reduce glycolysis, thereby preserving glycogen availability (Cox et al., 2016). Early experiments in dogs and rats have shown that ketosis can stimulate muscle glucose uptake (Madison et al., 1964) as well as glycogen storage (Maizels et al., 1977). Furthermore, ketosis generated by either a chronic high-fat, low-carbohydrate diet (Phinney et al., 1983) or by KB intake (Cox et al., 2016) decreased glycogen breakdown both at rest (Cox et al., 2016) and during exercise (Phinney et al., 1983; Cox et al., 2016). On the other hand, inducing ketosis by adhering to a ketogenic diet did not affect glycogen repletion patterns after a 3-h run (Volek et al., 2016). Based on these data, we hypothesized that during recovery from high-intensity exercise, KE ingestion could promote glycogen resynthesis by reducing disposal of glucose units via glycolytic breakdown. Suppression of glycolysis increases glucose-6-phosphate availability, which is a potent stimulator of the glycogen synthase enzyme (Jensen and Richter, 2012). However, contrary to our hypothesis, we found KE not to affect glycogen resynthesis during the 5 h recovery period. These data are in apparent contradiction with the results of another study in which KE increased muscle glycogen resynthesis following glycogen-depleting exercise (Holdsworth et al., 2016). In this study, however, CHO was delivered intravenously during a 2 h hyperglycaemic clamp. Under those conditions, KE ingestion in combination with carbohydrates resulted in 2-fold higher plasma insulin levels compared to carbohydrates alone. This was conceivably due to stimulation of pancreatic insulin secretion by high circulating plasma KB (Balasse et al., 1970; Beylot et al., 1986). Conversely, high-rate carbohydrate intake in conjunction with proteins in the conditions of the current study resulted in substantially higher plasma insulin levels than in the aforementioned studies due to the synergistic action of glucose and amino acids on pancreatic insulin release (Floyd et al., 1970). In addition, plasma insulin was identical between KE and PL (Figure 3A).
In the current study, subjects ingested 1.5 g·kg−1 BW KE, concentrations that have been reported previously to efficiently enhance blood ketone levels over longer time periods (Clarke et al., 2012b; Cox et al., 2016). In agreement with these studies, intermittent ingestion of KE with a protein-carbohydrate mixture resulted in blood βHB levels of ~5 mM but also resulted in gastrointestinal discomfort in half of the subjects. Gastrointestinal issues have been reported before under nutritional conditions not involving concomitant high-rate protein and carbohydrate intake (Clarke et al., 2012b). Although we cannot exclude that GI discomfort might have affected our experiment, biochemical measurements and circulating concentrations of βHB, glucose, and insulin were identical between subjects experiencing GI distress and subjects with adequate GI comfort. Future studies will need to elucidate more optimal dosing strategies to maintain the effect of KE on muscle protein synthesis in the absence of GI complaints.
In conclusion, while ingestion of KE did not increase muscle glycogen resynthesis, it markedly increased mTORC1 signaling following high-intensity exercise. Using C2C12 myotubes, we confirmed that KB potentiated leucine mediated mTORC1 activation and that this resulted into higher leucine-mediated protein ynthesis. Our results indicate that KE enhances the anabolic response to both exercise and protein ingestion but does not affect muscle glycogen synthesis, suggesting that it might become an attractive nutritional strategy to increase training-induced muscle remodeling and adaptation.