Reagents

Alpha modification of Eagle’s medium (AMEM, #310-010-CL), phosphate-buffered saline (PBS, #311-010-CL), trypsin (#325-043-CL), and antibiotic–antimycotic (ab-am, #15240-062) preparations were purchased from Wisent (St Bruno, Quebec, Canada). Foetal bovine serum (FBS, #12483-020), horse serum (HS, #26050088), Lipofectamine RNAiMAX (#13778-150), and Opti-MEM 1X Reduced Serum Medium (#31985070) were purchased from Thermo Fisher Canada (Burlington, Ontario Canada). Amino acid-free Roswell Park Memorial Institute (RPMI) 1640 medium (R8999-04A) was purchased from US Biologicals (Salem MA). L-leucine (#L8912), L-isoleucine (#I7403), L-valine (#V0513), sodium 4-methyl-2-oxovalerate (sodium salt of KIC, #K0629), 3-methyl-2-oxovaleric acid (sodium salt of KMV, #K7125), 3-methyl-2-oxobutyrate (sodium salt of KIV, #198994), protease (#P8340) and phosphatase (#P5726) inhibitor cocktails, anti-gamma tubulin antibody (#T6557), siRNA oligonucleotides, amino acid standard (#AAS-18), dimethyl sulfoxide (DMSO) (#D5879-100ML), O-phthalaldehyde (OPA, #P1378), 4,6-dimorpholino-N-(4-nitrophenyl)-1,3,5-triazin-2-amine (MHY1485, #SML0810), 1,2-diamino-4,5-methylenedioxybenzene (DMB, #66807), trypan blue dye (#T8154), and 3,6-dichlorobenzo[b]thiophene-2-carboxylic acid (BT2, #592682) were purchased from Sigma Aldrich (Oakville, Ontario, Canada). Rapamycin (#Rap004) was purchased from BioShop (Burlington, Ontario, Canada). Antibodies against phosphorylated (ph)-S6K1 (Thr389, #9234, 1:1000), ph-S6 (Ser235/6, #4858, 1:2000), ph-Akt (Ser473, #4060, 1:1000), ph-BCKDH-E1α (Ser293, #40368, 1:1000), ph-ACC (Ser 79, #3661, 1:1000) and ph-IRS-1 (Ser612, #3203, 1:1000), as well as horseradish peroxidase (HRP)-conjugated anti-rabbit (#7074, 1:10000) and anti-mouse (#7076, 1:10000) secondary antibodies were purchased from Cell Signaling Technology (Danvers, MA). Antibody against BCAT2 (#16417-1-AP) was purchased from ProteinTech (Rosemont, IL). Antibody against BDK (#PA5-31455) and Pierce BCA protein assay kit (#23225) were purchased from Thermo Fisher Canada (Burlington, Ontario, Canada). [³H]-2-deoxyglucose (#NET549) was purchased from Perkin Elmer (Markham, Ontario, Canada), U-¹⁴C labelled valine (ARC-0678) was purchased from American Radiolabeled Chemicals, while chemiluminescence substrate (#WBKLS0500) was from Millipore (Etobicoke, Ontario, Canada). L6 rat skeletal muscle myoblasts (#CRL-1458) were purchased from American Type Culture Collection (Manassas, VA).

Cell culture

L6 rat skeletal muscle cells were cultured in 10-cm plates with growth medium (GM: AMEM supplemented with 10% FBS and 1% antibiotic-antimycotic preparations). Cells were seeded (2 × 10⁵cells/well) in 6-well plates for western blot experiments or (10⁵ cells/well) in 12-well plates for glucose transport experiments. They were allowed to proliferate for 48 h or until they became 90–100% confluent. They were then shifted into the differentiation medium (DM: AMEM, 2% HS, 1% antibiotic-antimycotic preparations) and replenished with fresh DM every 48 h. Myotubes were used on day 5 (D5) of differentiation.

BCAA and BCKA treatment

Myotubes were treated with BCAA or BCKA for 30 min. They were then treated with or without 100 nM insulin and the treatment of BCAA/BCKA continued with insulin for 20 minutes. For KIC (Figs. 1–2 and 4–5), the concentration used was 200 μM. For leucine (Fig. 3), the concentration was 150 μM. For the BCKA treatment (Fig. 2), 200 μM was used, made up of 76 μM of KIC, 62 μM of KMV and 62 μM of KIV. For the BCAA treatment (Fig. 3), 400 μM was used, made up of 175 μM of valine, 150 μM of leucine, and 75 μM isoleucine. We used 200 μM of KIC as this concentration induces insulin resistance in L6 myotubes,^{(Reference Mann and Adegoke6,Reference Moghei, Tavajohi-Fini, Beatty and Adegoke7)} which also prompted us to use a similar value for total BCKAs supplementation. We used 150 μM of leucine as this concentration reduces glucose transport in L6 myotubes the most compared to higher concentrations.^{(Reference Moghei, Tavajohi-Fini, Beatty and Adegoke7)} The values used for the individual BCAAs are similar to values that are observed in plasma.^{(Reference Blair, Neinast and Jang21,Reference Gawedzka, Grandys, Duda, Zapart-Bukowska, Zoladz and Majerczak26)}

Fig. 1. BDK depletion attenuates the suppressive effect of KIC on insulin-stimulated glucose transport and the activation of S6K1/IRS-1. L6 myotubes were transfected with control (SCR) or BDK siRNA oligonucleotides. Forty-eight h later, cells were starved in a medium lacking amino acids and serum for 3 h. They were then treated without (−KIC) or with 200 μM KIC (+KIC) for 30 min. After, cells were incubated with or without 100 nM insulin for 20 min. Proteins in lysates were immunoblotted against BDK (a). Glucose transport assay was performed (b). Proteins in lysates were also immunoblotted against ph-S6K1^Thr389 (c, d), ph-IRS-1^Ser612 (c, e), ph-Akt^Ser473 (c, f). Proteins for western blot were normalised to γ-tubulin as the loading control. Glucose transport was normalised to the no insulin (–insulin) group in the SCR condition. n = 3–4 biological replicates with 3 technical replicates per experiment. Data are presented as Means ± SD * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001.

Fig. 2. BDK inhibition attenuates the suppression of glucose transport by BCKA. L6 myotubes were incubated for 3 h in 250 μM of BT2 in a starvation medium that lacked amino acids and serum. Incubation then continued in starvation medium with BT2 along with the addition of KIC (200 μM) or BCKA (total 200 μM: consisting of 76 μM of KIC and 62 μM for each of KMV and KIV) for 30 min. After, cells were incubated with or without 100 nM insulin for 20 min. Glucose transport assay (a) was then performed. Protein in lysates were immunoblotted against ph-BCKD^Ser293 (b, c). BCKD activity assay was performed (d). Protein in lysates were also immunoblotted against ph-S6K1^Thr389 (b, e), ph-IRS-1^Ser612 (b, f), ph-Akt^Ser473 (b, g), and ph-ACC^Ser79 (b, h). Proteins for western blot were normalised to γ-tubulin as the loading control. Glucose transport was normalised to the no insulin (–insulin) group in the vehicle (VEH) condition. n = 3 biological replicates with 3 technical replicates per experiment. Data are presented as Means ± SD. * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001.

Fig. 3. Inhibition of BDK did not modify the effect of BCAA on insulin-stimulated glucose transport. L6 myotubes were incubated for 3 h in 250 μM of BT2 in a starvation medium that lacked amino acids and serum. Incubation then continued in starvation medium with BT2 along with the addition of leucine (150 μM) or BCAA (total 400 μM: consisting of 175 μM of valine, 150 μM of leucine and 75 μM of isoleucine) for 30 min. After, cells were incubated with or without 100 nM insulin for 20 min. Glucose transport (a) and BCKD activity assays (b) were then performed. Proteins in lysates were immunoblotted against ph-S6K1^Thr389 (c, d), ph-IRS-1^Ser612 (c, e), ph-Akt^Ser473 (c, f), and ph-ACC^Ser79 (c, g). Proteins for western blot were normalised to γ-tubulin as the loading control. Glucose transport was normalised to the no insulin (–insulin) group in the VEH condition. n = 3 biological experiments with 3 technical replicates per experiment. Data are presented as Means ± SD. * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001.

Fig. 4. Suppression of mTORC1 is required for the effect of BT2 on glucose transport. mTORC1 was activated with the use of TSC-2 depletion and mTOR activator MHY1485. (a) Cells were transfected with control (purple) or TSC-2 (red) siRNA oligonucleotides. Forty-eight h later, they were treated with or without 250 μM of BT2 and with or without 10 μM of MHY1485 in serum- and amino acid-free medium for 3 h. Afterwards, the cells were supplemented without (−KIC) or with 200 μM KIC (+KIC) for 30 min followed by incubation with or without 100 nM insulin for 20 min. After treatments, proteins in lysates were immunoblotted against ph-S6^Ser235/6 (b, c) and ph-IRS-1^Ser612 (b, d). Proteins for western blot were normalized to γ-tubulin as the loading control. Glucose transport assay was performed (e). Glucose transport was normalised to the no insulin (–insulin) group in the VEH condition. Data are presented as Means ± SD; n = 3 biological replicates with 3 technical replicates per experiment. * P < 0.05, ** P < 0.01.

Fig. 5. Treatment with rapamycin impairs the effect of KIC-BT2 treatments on insulin-stimulated glucose transport. L6 myotubes were incubated for 3 h in 250 μM of BT2 in a starvation medium that lacked amino acids and serum. Incubation then continued in the medium with BT2 or vehicle that was also supplemented with or without KIC (200 μM) for 30 min and with or without 50 nM rapamycin. After, cells were incubated with or without 100 nM insulin for 20 min. Proteins in cell lysates were then immunoblotted against ph-S6K1^Thr389 (a, b) and ph-S6^Ser235/6 (a, c). Proteins for western blot were normalized to γ-tubulin as the loading control. Cells underwent a glucose transport assay (d). Glucose transport was normalised to the no insulin (–insulin) group in the VEH condition. n = 3 biological replicates with 3 technical replicates per experiment. Data are presented as Means ± SD. * P < 0.05, ** P < 0.01, ***, P < 0.001, **** P < 0.0001.

Fig. 6. Schematic of the relationship between BCAA/BCKA catabolism and regulation of glucose transport in myotubes. In the first panel, a simplified insulin signalling pathway is shown. In the second panel, with KIC supplementation, KIC is converted back to leucine by BCAT2, which then activates mTORC1/S6K1, leading to the phosphorylation of IRS-1^Ser612 and thus hindering IRS-1 signalling downstream. In the third panel, with BDK depletion or BT2 treatment, BCKD activity is increased leading to reduced leucine levels, and corresponding attenuation of mTORC1 activation and rescuing of insulin-stimulated glucose transport.

Gene silencing

On D3 of differentiation, myotubes were transfected with 50 nM scrambled (control, SCR) or BDK (sense 5’-CUAUGCAUGGCUUUGGCUU, anti-sense 5’-GAUACGUACCGAAACCGAA) or TSC-2 (sense 5’-GAGAUUGUUCUGUCCAUAA, anti-sense 5’-CUCUAACAAGACAGGUAUU). Where indicated, cells were treated with 50 nM scrambled or BCAT2 (sense 5′-GAGUGCAUCCGCCAGCUCA, anti-sense 5′-UGAGCUGGCGGAUGCACUC) siRNA oligonucleotides. Transfection was done using Lipofectamine RNAiMAX reagent according to the manufacturer’s instructions.

BT2 treatment

3,6-dichlorobenzo[b]thiophene-2-carboxylic acid (BT2) is an inhibitor of BDK, a protein kinase that is a negative regulator of BCKD.^{(Reference Neinast, Jang and Hui27,Reference Tso, Gui and Wu28)} A study that used a range between 160 and 750 μM of BT2 showed that 250 μM reduced ph-BCKD (Ser 293) the most in muscle cells.^{(Reference Biswas, Dao and Mercer5)} On D5 of differentiation, L6 myotubes were treated with 250 μM of BT2 in a starvation medium (free of amino acids and serum) for 3 h. Incubation then continued in starvation medium with BT2 or vehicle that was also supplemented with or without KIC (200 μM) for 30 min. After, incubation continued with or without BT2 and KIC. Cells were then incubated with or without 100 nM insulin for 20 min followed by glucose transport assay, BCKD activity assay, or harvesting for immunoblotting and high-performance liquid chromatography (HPLC) analyses. These treatments were done similarly in BDK-depleted cells (Fig. S4).

mTORC1 activation/inhibition experiments

To examine the contribution of mTORC1 to KIC-induced suppression of insulin-stimulated glucose transport, we initially used TSC-2-depleted cells to activate mTORC1. However, since insulin already diminishes TSC-2 inhibition of mTORC1, the increase in mTORC1 activation from TSC-2 knock-down was only marginally greater in insulin-treated cells (data not shown). Therefore, to effectively activate mTORC1 signalling under the conditions being tested, we combined TSC-2 depletion with treatment of the cells with 4,6-dimorpholino-N-(4-nitrophenyl)-1,3,5-triazin-2-amine (MHY1485), an mTOR activator.^{(Reference Choi, Park and Park29)} Myotubes depleted of TSC-2 were treated with vehicle or 10 μM of MHY1485 in a starvation medium for 3 h (Fig. 4a). Incubation then continued in starvation medium with MHY1485 that was also supplemented with or without KIC (200 μM) for 30 min. Myotubes were then incubated with or without 100 nM insulin for 20 min followed by glucose transport assay, or cells were harvested and processed for immunoblotting.

To suppress mTORC1 activity, we used rapamycin as an mTORC1 inhibitor,^{(Reference Lamming30)} as done previously in our work.^{(Reference Moghei, Tavajohi-Fini, Beatty and Adegoke7)} First, cells were starved with or without BT2 (250 μM) for 3 h. They were then treated with or without BT2 in the presence of KIC (200 μM), and with or without rapamycin (50 nM) for 30 min. Cells were then incubated with or without 100 nM insulin for 20 min with or without BT2, KIC, and rapamycin. They were then used for glucose transport assay or harvested for immunoblotting.

Glucose transport

Following treatments, myotubes cultured in 12-well plates were washed twice with HEPES [4-(2-hydroxy-ethyl)piperazine-1-ethanesulfonic acid]-buffered saline]. They were then incubated in 300 μl of transport solution (HEPES buffer, pH 8, 10 μM 2-deoxyglucose, 0.5 μCi/ml [³H]-2-deoxyglucose) for 5 min at 37°C and 2-deoxyglucose transport was performed as described previously.^{(Reference Moghei, Tavajohi-Fini, Beatty and Adegoke7,Reference Somwar, Sweeney, Ramlal and Klip31)} Following the 5-min incubation, plates were placed on ice and cells were washed with ice-cold saline 3 times. They were then lysed with 1 ml of cold 0.05 M NaOH. An aliquot of the lysate was used to determine protein concentration while another aliquot was counted. Glucose transport was calculated and expressed as pmol of deoxyglucose transported per μg protein as described before.^{(Reference Mann and Adegoke6)} These values were then normalised to the no-insulin group.

Amino acid concentrations

Amino acid concentrations were determined as previously described.^{(Reference Mann and Adegoke6)} Briefly, following treatments, myotubes were washed 2× in PBS, and then harvested with 10% trichloroacetic acid. Lysates were centrifuged at 10000 g for 15 min. Supernatant containing free amino acids were neutralised in a 1:2:1:8 ratio (sample: potassium phosphate buffer: 0.1 N hydrochloric acid: HPLC-grade water, respectively). Neutralised samples were pre-column derivatized with a 1:1 ratio of sample to o-Phthalaldehyde (29.28 mM). They were then injected into a YMC-Triart C18 column (C18, 1.9 μm, 75 × 3.0 mm) fitted onto an ultra-high-pressure liquid chromatography (UHPLC) system that was connected to a fluorescence detector (excitation: 340 nm; emission: 455 nm). Amino acids were eluted with a gradient solution derived from 20 mM potassium phosphate buffer (6.5 pH) (mobile phase A) and a solution made from 45% acetonitrile, 40% methanol and 15% HPLC-grade water (mobile phase B) at a flow rate of 0.8 ml/min. We used a gradient of 5–100% of mobile phase B over 21 min. Amino acid concentrations were calculated using amino acid standard curves and were normalised to total protein that was measured using the Pierce bicinchoninic acid (BCA) protein assay kit.

BCKA concentrations

Protocol was adapted from a previous study.^{(Reference Fujiwara, Hattori, Ito, Funatsu and Tsunoda32)} Following treatments, myotubes were washed 2× in PBS, then harvested with a lysis buffer (1 mM ethylenediaminetetraacetic acid (EDTA), 2% sodium dodecyl sulfate (SDS), 25 mM Tris-HCl pH 7.5, 10 μl/ml protease inhibitor cocktail, 10 μl/ml phosphatase inhibitor cocktail, 1 mM dithiothreitol (DTT)). Lysates were centrifuged at 10000 g for 10 min. The resulting supernatant was diluted in a 1:2:1:8 ratio (sample: potassium phosphate buffer: 0.1 N hydrochloric acid: HPLC-grade water, respectively). Diluted samples were treated with a 1:1 ratio of a 1,2-diamino-4,5-methylenedioxybenzene (DMB) solution (13.32 mM), sodium sulfite (38.88 mM), 2-mercaptethanol (1 M), HCl (0.696 M) in ddH₂O). Once samples were treated with DMB, this solution was heated at 85^oC for 45 minutes and then cooled on ice for at least 5 min. Samples and identically processed standards were injected into an Inertsil ODS-4 column (2 μM, 100 × 2.1 mm; GL Sciences, Torrance, CA, USA) fitted onto an ultra-high-pressure liquid chromatography system (Nexera X2, Shimadzu, Kyoto, Japan) that was connected to a fluorescence detector (Shimadzu, Kyoto, Japan; excitation: 367 nm; emission: 446 nm). Mobile phases were: (A) HPLC-grade methanol/ddH₂O (30/70, v/v), and (B) HPLC-grade methanol. Gradient elution was performed as follows: 0 min 0% B, 3.33 min 0%B, 5 min 50%B, 17.34 min 50%B. The flow rate was 0.2 ml/min, and the column temperature was maintained at 40^oC. BCKA concentrations were normalised to total protein as described above for the amino acids.

BCKD activity assay

BCKD catalyses the irreversible oxidative decarboxylation of the BCKA. Following treatments, each well was treated with 200 μl of starvation media and 856 μL of Krebs Ringer Buffer (0.018 M NaHPO₄, 0.68% (w/v) NaCl, 0.045% (w/v) KCl, and 0.03% (w/v) MgSO₄) that was supplemented with 1 mg of thiamine hydrochloride. Then, 61 μl of the reaction mixture (unlabelled valine (18.5 mM) and 1-¹⁴C labelled valine (3.7 μM) in PBS) was added to this. Wicks soaked with 60 μl of 2 M NaOH were taped hovering above each well. CO₂ released from the oxidative decarboxylation reaction was captured in 2 M NaOH-soaked filter paper wicks. This radiolabelled bicarbonate on the filter paper wick was transferred into a 20 ml scintillation vial containing 3.5 ml of scintillation fluid and counted. BCKD activity was calculated by dividing the radioactivity counts by the specific activity of valine in the reaction mixture (pmol). This value was then divided by the protein concentration in each well to obtain pmol/μg protein. These values were normalised to the vehicle control group.

Western blot

Following treatments, cells were processed as described previously.^{(Reference Jeganathan, Abdullahi, Zargar, Maeda, Riddell and Adegoke33,Reference Zargar, Moreira and Samimi-Seisan34)} Briefly, cells were harvested in lysis buffer (1 mM EDTA, 2% SDS, 25 mM Tris-HCl, pH 7.5, 1 mM DTT, and 10 μl/ml of each of protease inhibitor and phosphatase inhibitor cocktails). Proteins were separated on 10% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) followed by transfer onto polyvinylidene difluoride (PVDF) membranes (0.2 μm pore size). Transfer efficiency was checked with a Ponceau S incubation. This dye was then washed off with three 5-minute washes of Tris Buffered Saline with Tween (TBST). Next, membranes were incubated for one hour in 5% non-fat milk in TBST at room temperature to block non-specific antigen binding. Subsequently, they were quickly washed 3 times, 5 minutes each with TBST at room temperature and then incubated overnight at 4°C with the primary antibody of interest. Primary antibodies used for western blot include ph-S6K1, ph-IRS-1, ph-Akt, ph-S6, ph-ACC, ph-BCKD, BDK, and BCAT2. Proteins for western blot were normalised to γ-tubulin as the reference protein. Following the overnight incubation in primary antibody, membranes were washed 3 times for 5 minutes each with TBST and were incubated in a secondary antibody for three hours at room temperature. Secondary antibodies were diluted into a 5% milk with TBST solution before incubation with the membranes. HRP-conjugated secondary antibodies (anti-rabbit or anti-mouse) were used at 1:10000 dilution in 5% non-fat milk in TBST. Subsequently, membranes were washed 3 times for 5 minutes each with TBST before HRP chemiluminescence substrate was applied. Bio-Rad ChemiDoc XRS+ was used for signal visualisation and the images were quantified with Image Lab software (version 8).

Data presentation and statistical analysis

Glucose transport data was normalised to the no-insulin group. BCKD activity data are normalised to the vehicle control group. Data are presented as means ± SD. One-way analysis of variance was used and Tukey’s post-hoc tests were done to measure statistically significant differences among means. Significance was determined as P < 0.05. Statistical analyses were performed using GraphPad Prism software (GraphPad, Boston, MA).