Eleven healthy (body mass: 83.3 ± 11.7 kg; percent lean mass: 86.1 ± 4.1%) young (20.1 ± 1.3 years) men who regularly participated in resistance training of the lower body (≥2/week for >1 years) volunteered to participate in the present study. A priori, we conducted a power calculation (GPower v3 software) of appropriate sample size based on our previous published data (Witard et al., 2014) that measured, on average, a 37% higher post-exercise response of myofibrillar-MPS to 20 g of ingested whey protein compared with 0 g using the same tracer technique. The BCAA trial in the present study contained a similar quantity of BCAAs as the 20 g bolus of whey protein administered in our previous study (Witard et al., 2014; 5.6 vs. 4.8 g). By setting statistical power (1-ß err prob) at 0.8, α error probability at 0.05 and effect size (Cohen's d) at 1.4 (based on our previous data; Witard et al., 2014), our power calculation revealed a minimum sample size of 10 participants (using a crossover research design) would be necessary to detect a statistical difference in myofibrillar-MPS between BCAA and placebo trials. Due to illness, only 10 participants completed both experimental trials, therefore all data are shown for 10 participants.

The design, purpose, and potential risks associated with the study were explained to participants before obtaining written informed consent. Exclusion criteria were metabolic disorders, food intolerances/allergies, allergies to local anesthetic, current participation in other clinical trials, blood donations within three months of the initial screening visit, and the prescription of medication or consumption of nutritional/dietary supplements suggested to affect protein metabolism. The National Research Ethics Service board (Warwickshire, Birmingham, UK) approved all procedures and all were in accordance with the Helsinki Declaration of 1975 as revised in 1983. This trial is registered as a clinical trial (registration number: ISRCTN98737111).

Participants reported to the laboratory on five separate occasions, including two experimental trials that were separated by approximately three weeks. During the initial visit, body composition was assessed using dual-energy x-ray absorptiometry (DEXA; Hologic Discovery W, Hologic Inc., Bedford, Massachusetts, USA) and one repetition maximum (1RM) was predicted for each leg individually on both leg press and leg extension. Approximately one week later, participants returned to the laboratory to confirm their 1RM for each leg. Three days later, participants performed their first blinded trial in which they consumed either a BCAA containing drink (BCAA; GlaxoSmithKline, Brentford, UK) or placebo (PLA) drink (GlaxoSmithKline, Brentford, UK; Table 1). BCAA contained 5.6 g of BCAAs which is equivalent to the typical BCAA content of 20 g whey protein. Participants performed a unilateral bout of resistance exercise prior to consuming the test drink in each trial. Myofibrillar-MPS was determined by measuring the incorporation of L-[ring-¹³C₆] phenylalanine into myofibrillar protein during a primed continuous infusion. The complete testing procedure was repeated on the contralateral leg approximately three weeks after the first trial. Trial and exercised leg order were counter-balanced and randomized.

During the initial visit, body composition (total and segmental) was assessed using DEXA. Prior the experimental trial, participants completed a questionnaire of food preferences and a 3-day diet diary that represented their habitual daily intakes. The average energy intake (3,392 ± 1,069 kcal) and macronutrient composition [protein: 2.2 ± 1.0 g·kgBM⁻¹·d⁻¹; carbohydrate: 4.7 ± 1.6 g·kgBM⁻¹ ·d⁻¹; fat: 1.4 ± 0.4 g·kgBM⁻¹·d⁻¹] from the 3 day diet diary was used to calculate the participant's diet before the experimental trial. Food parcels, which matched each participant's habitual energy and macronutrient intakes, were supplied for 48 h before the experimental trial. Participants were instructed to consume only food and drink sources provided by investigators and to consume their final meal no later than 22:00 on the evening before the experimental protocol. Analyses of the diet diary and food prescription were performed by using a commercially available software program (Wisp v3; Tinuviel software). All participants were instructed to refrain from physical exercise for 48 h before the experimental trial.

A schematic diagram of the experimental protocol is displayed in Figure 1. Participants reported to the laboratory at ~06:15 following an overnight fast and measurements of height and weight were collected. A cannula was inserted into a forearm vein of one arm, and a resting blood sample was collected. Thereafter, participants were fed a standardized energy-rich (7 ± 1 kcal·kgBM⁻¹), high protein (30 ± 1% of energy from protein, 50 ± 1% of energy from carbohydrate, and 20 ± 1% of energy from fat) breakfast. After breakfast a primed, continuous infusion (prime: ~88 μmol·kg⁻¹; infusion rate: ~0.227 μmol·kgBM⁻¹·min⁻¹) of ¹⁵N₂ urea was started and continued throughout the protocol. Subjects rested for ~75 min before a primed, continuous (2.0 μmol·kg⁻¹; infusion ~0.05 μmol·kgBM⁻¹.min⁻¹) infusion of L-ring ¹³C₆ phenylalanine (both Cambridge Isotope Laboratories Inc., Andover, Massachusetts, USA) was started.

Approximately 105 min later, participants performed a single bout of unilateral leg resistance exercise which lasted ~25 min. Participants completed a warm up on the leg press machine (Cybex VR3, Cybex International, Medway, Massachusetts, USA), which consisted of 12 repetitions at 40% 1RM, 10 repetitions at 50% 1RM, eight repetitions at 60% 1RM, and two repetitions at 70% 1RM. The warm up was followed by a 2 min rest period. The resistance exercise protocol consisted of four sets of 10 repetitions at 70% 1RM on the leg press machine and four sets of 10 repetitions at 75% 1RM on the leg extension machine (Cybex VR3, Cybex International, Medway, Massachusetts, USA). Participants rested for 2 min between each set and were free to consume water ad libitum. If a participant could not complete a full set, the load was lowered by 4.5 kg. Rating of perceived exertion (RPE) using the modified Borg scale (Borg, 1973) was measured after each set. Immediately (~5 min) after exercise (t = 0 min), an arterialized blood sample and a muscle biopsy were collected followed by ingestion of the test drink. Blood samples also were collected at multiple time points (t = −240, −145, −85, −25, 0, 15, 30, 45, 60, 75, 90, 120, 150, 180, 240 min) before and after drink ingestion. Two further muscle biopsies were collected 1 and 4 h after drink ingestion.

Muscle biopsies were obtained from the vastus lateralis of the exercised leg under local anesthesia (1% lidocaine) using 5 mm Bergstrom needles with suction. Different incisions (~1 cm apart, proximally from previous site) were used for each biopsy in an attempt to minimize the impact of local inflammation from the previous biopsy sample. Biopsy samples were immediately rinsed, blotted of excess blood, removed of visible fat and connective tissue, and divided into aliquots, before being frozen in liquid nitrogen, and stored at –80°C until later analysis. Muscle samples were analyzed for enrichment of L-[ring-¹³C₆] phenylalanine in the intracellular pool and bound myofibrillar protein fractions, Furthermore, muscle was analyzed to measure the phosphorylation status of mTORC1-related signaling proteins.

Muscle tissue was analyzed for enrichment of L-[ring-¹³C₆] phenylalanine in the myofibrillar protein fraction. Myofibrillar proteins were isolated from ~30 mg tissue as previously described (Moore et al., 2009). Briefly muscle was snipped in ice-cold homogenizing buffer (50 mM Tris-HCL, 1 mM EDTA, 1 mM EGTA, 10 mM β-glycerophosphate, and 50 mM sodium fluoride). The homogenate was shaken for 10 min prior to being centrifuged at 1,000 g for 10 min at 4°C. The pellet was then re-suspended in homogenization buffer, before being shaken and centrifuged as described above. The myofibrillar fraction was separated from any collagen by dissolving the pellet in 0.3 M NaCl and heating for 30 min at 37°C. The proteins were then precipitated by combining the supernatant with 1 M PCA before being centrifuged for 20 min at 3500 rpm at 4°C. The pellet was then washed with 70% ethanol. The remaining myofibrillar pellet was hydrolyzed overnight at 110°C in 0.1 M HCl/Dowex 50W-X8 100–200 (Bio-Rad laboratories Inc., USA) and the liberated amino acids purified on cation-exchange columns (Bio-Rad laboratories Inc., USA). Amino acids were then converted to their N-acetyl-n-propyl ester derivative and phenylalanine labeling was determined by gas-chromatography-combustion-isotope ratio mass spectrometry (GC-C-IRMS, Delta-plus XL, Thermofinnigan, Hemel Hempstead, UK). Ions 44/45 were monitored for unlabeled and labeled CO₂, respectively.

Intracellular amino acids were liberated from ~20 mg of muscle. The frozen tissue was powdered under liquid nitrogen using a mortar and pestle and 500 μL of 1 M perchloric acid (PCA) was added. The mixture was centrifuged at 10,000 g for 10 min. The supernatant was then neutralized with 2 M potassium hydroxide and 0.2 M PCA and combined with 20 μL of urease for removal of urea. The free amino acids from the intracellular pool were purified on cation-exchange columns as described above. Intracellular amino acids were converted to their MTBSTFA derivative and ¹³C₆ phenylalanine enrichment was determined by monitoring at ions 234/240 using gas chromatography mass spectrometry (GCMS; model 5973; Hewlett Packard, Palo Alto, CA).

Muscle tissue (20–30 mg) was powdered on dry ice and then homogenized in lysis buffer (50 mM Tris pH 7.5; 250 mM Sucrose; 1 mM EDTA; 1 mM EGTA; 1% Triton X-100; 1 mM NaVO₄; 50 mM NaF; 0.50% PIC) using a hand-held homogenizer (PRO200, UK). Samples were placed on a shaker for 1 h at 4°C, before being centrifuged for 5 min at 6,000 g. The supernatant was then used for determination of protein. A DC protein assay (Bio Rad, Hertfordshire, UK) was used for determining protein concentration. Equal amounts of protein were then boiled in Laemmli sample buffer (250 mM Tris-HCl, pH 6.8; 2% SDS; 10% glycerol; 0.01% bromophenol blue; 5% β-mercaptoethanol) and separated on SDS polyacrylamide gels (10–15%) for ~1 h at 58 mA. Proteins were then transferred to a Protran nitrocellulose membrane (Whatman, Dassel, Germany) at 100 V for 1 h. Membranes were blocked using milk solution and then incubated overnight at 4°C with the appropriate primary antibody. The primary antibodies used were: S6K1^Thr389 (Cell signaling #9234); Akt^Ser473 (Cell signaling #3787); 4E-BP1^Thr37/45 (Santa Cruz SC6025); and PRAS40^Thr246 (Cell Signaling #2610). The following morning the membrane was rinsed in wash buffer (TBS with 0.1% Tween-20) before being incubated for 1 h at room temperature with the appropriate secondary antibody; either horseradish (HRP)-linked anti-mouse IgG (New England Biolabs, 7072;1:1,000) or anti-rabbit IgG (New England Biolabs, 7074; 1:1,000) diluted in wash buffer. The membrane was then cleared of the antibody using wash buffer. Antibody binding was detected using enhanced chemiluminescence (Millipore, Billerica, MA). Imaging and band quantification were carried out using a Chemi Genius Bioimaging Gel Doc System (Syngene, Cambridge, UK). Blots were stripped in Restore™ Western Blot Stripping Buffer (Pierce, Rockford, IL) for 20 min and, re-blocked in milk solution and the respective total protein antibodies [S6K1 (Santa Cruz #sc-230); Akt (Cell signaling #9272); 4E-BP1 (Santa Cruz #sc-4251); and PRAS40 (Cell Signaling #2691)] were used to assure equal loading.

Blood was collected in lithium heparin, EDTA-containing and serum separator tubes and centrifuged at 3,500 rpm for 15 min at 4°C. Aliquots of plasma and serum were then frozen at −80°C until later analysis. Plasma glucose and urea concentrations were analyzed using an instrumentation laboratory automated blood metabolite analyzer (Instrumentation Laboratory 650, Instrumentation Laboratory, Cheshire, UK). Serum insulin concentrations were measured using a commercially available ELISA (DRG Diagnostics, USA) following manufacturer's instructions.

The ¹³C₆ enrichments of phenylalanine and tyrosine were determined by GCMS by monitoring ions 234 and 240 for unlabeled and labeled phenylalanine and ions 466 and 472 for unlabeled and labeled tyrosine. Once thawed, plasma samples were mixed with diluted acetic acid and purified using a cation-exchange column (Bio-Rad laboratories Inc., USA). The amino acids were then converted to their N-tert-butyldimethyl-silyl-N-methyltrifluoracetamide (TBDMS) derivative.

Simultaneously, concentrations of phenylalanine, leucine, threonine, isoleucine, and valine were determined using an internal standard method (Tipton et al., 1999b, 2001). The selected amino acids were chosen to monitor blood concentrations of essential, non-essential, and the BCAAs. Briefly, plasma samples were weighed to obtain ~300 μL, 30 μL of an internal standard containing U-¹³C₉-¹⁵N phenylalanine (ions 336/346), U-¹³C₆ leucine, and isoleucine (ions 302/308), U-¹³C₄-¹⁵N threonine (ions 405/409), and U-¹³C₅ valine (ions 288/293) was then added and weighed. Since the weight of both sample and internal standard was known, it was possible to calculate a tracer-to-tracee ratio. Since the amino acid concentrations of the internal standard were known, it was possible to convert the tracer-to-tracee ratio into a concentration of each amino acid in plasma.

To determine ¹⁵N₂ urea enrichments, 10 μL of plasma was mixed with 120 μL ethanol. Samples were then left in the fridge for 30 min prior to centrifugation at 13,000 rpm for 20 min at 4°C. The supernatant was then removed and dried under nitrogen. TBDMS and acetonitrile were added to the dried sample prior to heating at 90°C for 90 min. Samples were then run on GCMS and ions 231 and 233 were monitored.

Total area under the curve (tAUC) for serum insulin concentrations, phenylalanine Ra, and phenylalanine oxidation rates were calculated using Graphpad Prism V5.0 (Graphpad software incorporation, La Jolle, California, USA). tAUC-values were calculated from drink ingestion to final blood sample and a baseline y-axis value of zero for each was used.

A two-way repeated measures ANOVA was used to analyse blood glucose, insulin, amino acid, and urea concentrations, whole body urea production rates, phenylalanine Ra, phenylalanine oxidation rates, and phenylalanine/tyrosine enrichments (within-subject factor: time point; between-subject factor: trial). Exercise variables, myofibrillar-MPS, and tAUC of serum insulin concentrations, and phenylalanine kinetics for the post-exercise period were analyzed using a paired samples one-tailed t-test. Western blot data were expressed as percentage change from t = 0 (immediately post-exercise muscle biopsy) and data were analyzed using a two-way repeated measures ANOVA. Given the large variability associated with phosphorylation measurements of anabolic cell signaling, a priori we decided to explore differences over time for each trial separately. Significance for all analyses was set at P < 0.05. Where significance was detected, a LSD correction was used in post-hoc analysis. All statistical tests were completed using statistical package for social sciences version 23.0 (IBM, Chicago, Illinois, USA). All values are presented as means ± SEM, unless otherwise stated.

The lean tissue mass of the exercised leg, measured prior to each experimental protocol, was similar between trials (P > 0.05). The total weight lifted throughout the exercise protocol (including warm-up) was similar between BCAA (11,055 ± 837 kg) and PLA (11,089 ± 913 kg) trials (P = 0.88) and participants reported RPE scores 19 in all exercise trials.

Plasma glucose concentrations did not change over time (P = 0.093), were similar between trials (P = 0.365; data not shown) and no interaction effect was observed (P = 0.548). In both trials, serum insulin concentrations markedly (264%) increased (P < 0.01) in response to breakfast, before returning to baseline 1 h after exercise and remained stable over the remainder of the recovery period (data not shown). There were no differences in insulin concentrations between BCAA and PLA trials (P = 0.48) and no interaction effect was observed (P = 0.36). The insulin tAUC over 4 h following drink ingestion was similar (P = 0.70) between BCAA (50 ± 17 μIU·mL^{−1 *}240 min) and PLA (53 ± 19 μIU·mL^{−1 *}240 min) trials.

A significant time effect was observed for plasma concentrations of leucine (P < 0.001), isoleucine (P < 0.001), and valine (P < 0.001; Figure 2). The concentrations of all branched chain amino acids were greater in BCAA than PLA (P < 0.001). Peak leucine (624 ± 41 μM), isoleucine (339 ± 23 μM), and valine (753 ± 45 μM) concentrations were observed at 0.5 h-post drink ingestion in BCAA.percentage increase from baseline of 300 ± 96% (leucine), 300 ± 88% (isoleucine), and 144 ± 59% (valine). A significant interaction effect for plasma branched chain amino acid concentrations was observed whereby over the entire measurement period after drink ingestion, leucine (P < 0.001), and valine (P ≤ 0.002) concentrations were greater in BCAA than PLA. Over a 3.5 h period post drink ingestion, isoleucine levels were higher in BCAA (P ≤ 0.001).

Phenylalanine and threonine concentrations declined to below baseline values after drink ingestion in both trials (P < 0.001). There were no difference in threonine concentrations between trials (P = 0.207) and no interaction effect was observed (P = 0.056). Phenylalanine concentrations (main effect of trial; P = 0.003) were significantly higher in BCAA than PLA at 15 min (83 ± 5 vs. 75 ± 8 μM, P = 0.018), but significantly lower in BCAA than PLA at 120 min (57 ± 9 vs. 67 ± 8 μM, P = 0.003), 150 min (58 ± 7 vs. 65 ± 9 μM, P = 0.007), and 240 min (63 ± 6 vs. 73 ± 13 μM, P = 0.021) post drink ingestion.

A decrease in plasma urea concentrations was observed in both trials during the latter stage (3–4 h) of exercise recovery (P < 0.026), with no differences between trials (P = 0.733; data not shown). Whole body urea production rates decreased from immediately post-exercise to the end of the 4 h recovery period (P < 0.001; data not shown), however there was no differences between trials (P = 0.685) and no interaction effect was observed (P = 0.611). The tAUC of urea production in response to post-exercise drink ingestion was similar between BCAA (2,314 ± 490 μmol·kgBM⁻¹) and PLA (2,272 ± 636 μmol·kgBM⁻¹ P = 0.694) trials.

Muscle intracellular and blood plasma ¹³C₆ phenylalanine enrichments were stable over the measured time period of tracer incorporation (0–4 h) in both trials (Figure 3). Plasma ¹³C₆ tyrosine enrichments increased over time during the experimental protocol (data not shown). Trial order had no effect on intracellular and plasma tracer enrichments.

Phenylalanine Ra, expressed as tAUC over the post drink period, was 6% lower in BCAA than PLA (P = 0.028). In addition, tAUC of whole body phenylalanine oxidation rate was 16% lower in BCAA compared with PLA (P = 0.006, Figure 4).

Signaling data are displayed in Figure 5. The phosphorylation of AKT^Ser473 increased over time (P = 0.025). However, there was no trial effect (P = 0.388). The phosphorylation of AKT^Ser473 increased at 1 h compared with 0 h in BCAA (P = 0.007). Further, phosphorylation of AKT^Ser473 was higher at 1 h compared with 4 h in BCAA (P = 0.015) but there were no differences between 0 and 4 h in BCAA (P = 0.903). There were no differences between 0 and 1 h (P = 0.056) or 4 h (P = 0.174) and 1 and 4 h (P = 0.085) in PLA.

The phosphorylation of PRAS40^Thr246 increased over time (P = 0.022). However, there was no trial effect (P = 0.281). The phosphorylation of PRAS40^Thr246 increased ~12-fold at 1 h compared with 0 h in BCAA (P = 0.037). Further phosphorylation of PRAS40^Thr246 was higher at 1 h compared with 4 h in BCAA (P = 0.047) but there were no differences between 0 and 4 h in BCAA (P = 0.102). There were no differences between 0 and 1 h (P = 0.050) or 4 h (P = 0.101) and 1 and 4 h (P = 0.112) in PLA.

The phosphorylation of S6K1^Thr389 increased over time (P = 0.021). However, there was no trial effect (P = 0.087). The phosphorylation of S6K1^Thr389 increased ~6-fold at 1 h compared with 0 h in BCAA (P = 0.017). Further, phosphorylation of S6K1^Thr389 was higher at 1 h compared with 4 h in BCAA (P = 0.014) but there were no differences between 1 and 4 h in BCAA (P = 0.690). There were no differences between 0 and 1 h (P = 0.120) or 4 h (P = 0.05)0 and 1 and 4 h (P = 0.795) in PLA.

The phosphorylation of 4EBP1^Thr37/45 remained constant over time (P = 0.115) and there were no trial effect (P = 0.566).

Myofibrillar FSR was ~22% higher (P = 0.012) in BCAA than PLA over the 4 h period following drink ingestion (Figure 6).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.