D-glucose (Glc), L-glutamine (Gln), L-glutamic acid (Glu), L-leucine (Leu), L-valine (Val), β-hydroxy butyric acid (βOHB), and FA-free bovine serum albumin (BSA) were purchased from FUJIFILM Wako Pure Chemicals (Osaka, Japan). Palmitic acid (PA), oleic acid (OA), and 3-(4,5-dimethylthiazol-2-y1)-2,5-diphenyltetrazoli-umbromide (MTT) were purchased from Nacalai Tesque (Kyoto, Japan). Sodium lactate (LA) and dimethyl α-ketoglutarate (αKG; an esterified form of α-ketoglutarate) were purchased from Otsuka Pharmaceutical (Tokyo, Japan) and Combi-Blocks (San Diego, CA, USA), respectively. All chemicals, except PA and OA, were dissolved in deionized distilled water.

The cells were treated with fatty acids (FAs) using a previously described method, wherein the fatty acids were individually complexed with BSA in phosphate-buffered saline (PBS) (22). Briefly, FAs (PA or OA) were added to 100 mM NaOH solution and dissolved in a heat-block at 75°C for 30 min to obtain the FA solution. The prepared FA solution was added to 10% (w/v) FA-free BSA in PBS at a ratio of 1:9 (v/v) to obtain a 10 mM FA solution. This solution was added to the medium at a final concentration of 0.1 or 0.2 mM FA. A mixture of 100 mM NaOH and 10% BSA/PBS (1:9) was used as the vehicle control. Each prepared solution was heated up to 55°C before being added to the cells.

Mouse C2C12 myoblast cell line (RCB0987) was purchased from the RIKEN BioResource Research Center (Tsukuba, Japan). The cells were cultured in Dulbecco's modified Eagle's medium (DMEM) with 5.5 mM glucose (041-29775; FUJIFILM Wako Pure Chemicals) supplemented with 10% fetal bovine serum at 37°C in a humidified atmosphere with 5% CO₂. To induce differentiation of myoblasts into myotubes, C2C12 cells at 80% confluency were maintained in the differentiation medium (DMEM containing 2% horse serum; Thermo Fisher Scientific, Tokyo, Japan) for 6 days. For starvation, the differentiation medium was replaced with ‘starvation medium’-DMEM lacking glucose, glutamine, and pyruvate (A14430; Thermo Fisher Scientific)- and the cells were incubated for the specified durations before harvesting. The control cells were incubated in regular DMEM without serum for the same durations. For some experiments, a specific nutrient was added to the starvation medium at concentrations within the medium to high physiological range, which are typically used in cell experiments. The concentrations of each nutrient used in this study are shown in Table I.

Cell viability was assessed using the MTT assay. The tetrazolium salt MTT is metabolized to insoluble purple formazan crystals in the mitochondria, which can be quantified upon solubilization using a spectrophotometer. The incubation medium was replaced with a medium containing MTT (0.5 mg/ml) 3 h before analysis. After the 3-h incubation period, the medium was replaced with dimethyl sulfoxide (250 µl/well in a 24-well plate) to dissolve the formazan crystals. The absorbance was measured at 570 nm using a microplate reader (Multiskan FC; Thermo Fisher Scientific JP, Tokyo, Japan).

Total RNA was isolated from cells using the RNAiso Plus reagent (TAKARA BIO, Shiga, Japan), and first-strand cDNA was synthesized using the ReverTra Ace qPCR RT Master Mix (TOYOBO, Osaka, Japan), according to the manufacturer's instructions. qPCR was performed using TB Green Premix Ex Taq II (TAKARA BIO) in 10-µl reactions on a StepOne Plus Real-Time PCR System (Thermo Fisher Scientific). The qPCR cycling conditions were as follows: One cycle of 30 sec at 95°C, followed by 40 cycles of 5 sec at 95°C, and finally, one cycle of 30 sec at 60°C. Gene expression was normalized to that of a standard housekeeping gene (β-actin) and analyzed with the 2^−ΔΔCq method (23). Experiments were performed in duplicate. The primer sequences are listed in Table SI.

Cells were washed and lysed with 50 µl of lysis buffer [1% Triton X-100, 0.45% sodium pyrophosphate, 100 mM NaF, 2 mM Na₃VO₄, 50 mM HEPES (pH 8.0), 147 mM NaCl, 1 mM EDTA, and a protease inhibitor mixture (cOmplete; Sigma-Aldrich, Tokyo, Japan)]. Then, the lysed cells were centrifuged at 13,000 × g, 4°C, for 15 min to obtain the supernatant, which was used as the cell lysate. Equal amounts of cell lysate proteins were loaded (10 µg per lane) onto a 15% acrylamide gel for microtubule-associated protein light-chain 3 (LC3) and onto a 10% acrylamide gel for other proteins, separated by SDS-PAGE electrophoresis, and transferred onto PVDF membranes (Hybond-P; GE Healthcare Life Science, Tokyo, Japan). The membranes were blocked with Blocking One-P (Nacalai Tesque) for 60 min at room temperature (22–24°C) and then incubated overnight (16–18 h) at 4°C with primary antibodies against each of the following proteins: LC3 (#2775), 70-kDa ribosomal protein S6 kinase (p70S6K; #9202), phosphorylated-p70S6K (#9205), AMP-activated protein kinase (AMPK; #2532), and phosphorylated-AMPK (#2535), all purchased from Cell Signaling (Danvers, MA, USA), or β-actin (#sc47778; Santa Cruz Biotechnology, Dallas, TX, USA). After washing with PBS containing 0.1% Tween 20 (TBS-T), the membranes were incubated with horseradish peroxidase-conjugated secondary antibodies for 1 h at room temperature (22–24°C). Specific proteins were detected by chemiluminescence using the ECL Select Western Blotting Detection Reagent (GE Healthcare). Images were captured and quantified using the iBright Imaging System (Thermo Fisher Scientific JP).

Differentiated C2C12 cells were fixed and permeabilized for 10 min at room temperature (22–24°C) using 1% Triton-X containing 4% paraformaldehyde in PBS. Then, the cells were incubated with MF-20 anti-MHC antibody (1:100; Hybridoma Bank, Iowa, IA, USA) for 90 min at room temperature, washed with PBS-T, and incubated for 1 h with fluorescein isothiocyanate-conjugated anti-mouse IgG (1:500; #115-095-062; Jackson ImmunoResearch, West Grove, PA, USA). For each condition, representative images of the cells were captured using a 10× objective lens on a fluorescence microscope (BZ-X700; Keyence, Osaka, Japan).

Cellular ATP production and glycolytic capacity were evaluated using the Glycolysis/OXPHOS Assay Kit (G270; Dojindo, Kumamoto, Japan), following the manufacturer's instructions. Differentiated myotubes were incubated without or with 2.5 µM oligomycin (an inhibitor of the mitochondrial respiratory chain) in a 96-well culture plate for 3 h. Then, ATP working solution was added to each well to assess ATP production. The plate was incubated for 10 min at 25°C, and relative ATP content was measured as luminescence intensity, using a multimode plate reader (EnSpire, Perkin Elmer Japan, Yokohama, Japan).

Data are expressed as the mean ± standard error of the mean. All statistical analyses were performed using the IBM SPSS Statistics for Windows (version 29; IBM Japan, Tokyo, Japan). The unpaired Student's t-test was used to identify significant differences between two groups, and one-way analysis of variance followed by Tukey's post-hoc test was performed to determine differences among three or more groups. P<0.05 was considered to indicate a statistically significant difference.

Mitochondrial dehydrogenases reduce MTT to formazan; thus, formazan production depends on the redox activity of the cell and reflects mitochondrial function (24). Consequently, the amount of formazan produced can be interpreted as an indicator of mitochondrial activity and can also be used to assess cell viability. In this study, the mitochondrial activity of C2C12 myotubes was suppressed by incubation in the starvation medium in a time-dependent manner (Fig. 1A). After 3, 8, and 24 h of starvation, the mitochondrial activity in myotubes decreased to approximately 60, 44, and 7%, respectively, when compared with that in untreated cells. The decrease in cell viability after 5 h of starvation was restored to levels comparable to those of control cells by replacing the medium with regular DMEM for 3 h (Fig. 1B). This indicates that short-term starvation of up to 5 h causes temporary suppression of mitochondrial activity rather than irreversible cell damage. In contrast, the decrease in cell viability after 8 h of starvation was not fully recovered by replacing the medium with regular DMEM (Fig. 1B). This irreversible decline in mitochondrial activity became more pronounced with longer starvation times (Fig. 1B), suggesting a transition from reversible suppression of mitochondrial respiration to irreversible cell damage due to prolonged starvation.

To evaluate the cellular metabolic response to nutrient starvation, we examined the expression levels of key metabolic genes using RT-qPCR. The names and roles of these genes are listed in Table SI. After 1 h of starvation, none of the genes tested showed a significant change in their expression levels (Fig. 2A). After 5 h of starvation, the gene expression of hexokinase 2 (Hk) was suppressed, whereas that of several genes involved in the uptake and degradation of fuel substrates, such as succinate-CoA ligase, subunit β (Sucla2), cluster of differentiation 36 (Cd36), medium-chain acyl-CoA dehydrogenase (Mcad), branched chain α-keto acid dehydrogenase E1α (Bckdha), and succinyl CoA 3-oxoacid CoA transferase (Scot), increased (Fig. 2B). In contrast, the expression of most metabolic genes was suppressed after 24 h of starvation, with the exception of Scot and carnitine palmitoyl transferase 1-b (Cpt1b), the latter showing a notable increase in expression (Fig. 2C).

The balance between protein degradation and synthesis is crucial for maintaining muscle mass and is regulated by the energy status of muscle tissue. Therefore, we investigated alterations in protein metabolism. After 1 h of starvation, the expression levels of atrogin-1 (Atg1) and muscle ring finger 1 (Murf1), which contribute to muscle proteolysis, remained unchanged. However, after 5 h of starvation, the expression levels of these genes increased significantly, accompanied by a trend toward AMPK activation (P=0.07) (Fig. 3A and B), whereas the LC3II/LC3I ratio, a representative autophagy marker, showed no change (Fig. 3B).

After 24 h of starvation, Atg1 and Murf1 expression remained elevated when compared with that in controls; however, the difference was less pronounced than that after 5 h of starvation (Fig. 3A). In contrast, the LC3II/LC3I ratio increased significantly, accompanied by a marked increase in AMPK activity (Fig. 3C). Regarding protein synthesis, the activity of p70S6K, which enhances the protein synthesis pathway, was significantly suppressed after both 5 and 24 h of starvation (Fig. 3B and C). In addition, the protein levels of AMPK, LC3, and p70S6K decreased after 24 h of starvation. When cultured cells are deprived of serum, a selective proteolytic pathway is activated for specific proteins, which applies to approximately 30% of cytosolic proteins (25,26). Therefore, these proteins may have been partially degraded by this pathway.

Next, we evaluated whether any of the nutrient substrates could improve the metabolic disturbances induced by nutrient starvation. MTT assay was utilized to assess the toxic effects of each nutrient at the concentrations used on mitochondrial function in myotubes maintained in a differentiation medium. PA treatment significantly impacted mitochondrial activity after 24 h of incubation (Fig. 4A); therefore, we excluded PA from further studies.

Glc replacement notably recovered the decrease in mitochondrial activity induced by 5 h of starvation, as expected. Starvation-induced metabolic inactivity in mitochondria was also attenuated by adding Gln, LA, or αKG to the starvation medium. However, other substrates did not achieve similar improvements at the tested concentrations (Fig. 4B). Even after 24 h of starvation, which caused suppression of redox activity to less than 5% of that of control cells, the addition of each of these four substrates (Glc, Gln, LA, or αKG) alleviated the decline in mitochondrial activity to varying extents (Fig. 4C).

We also examined the expression levels of genes whose expression changed after the cells were starved for 24 h. All four substrates tested (Glc, Gln, LA, or αKG) attenuated the decrease in the expression of the following genes in starved cells: Glucose transporter 4 (Glut4), oxoglutarate dehydrogenase (Ogdh), glutamate dehydrogenase 1 (Glud1), branched-chain amino transaminase 2 (Bcat2), Bckdha, lactate dehydrogenase (Ldh), and monocarboxylic acid transporter 1 (Mct1) (Fig. 5). Additionally, LA or αKG attenuated fasting-induced decrease in the expression of pyruvate kinase (Pk) and citrate synthase (Cs). Glc, Gln, or αKG attenuated that of Mcad, while they suppressed the fasting-induced increase in Cpt1b expression (Fig. 5).

We examined the effects of selected nutrients, including βOHB, which generates acetoacetic acid as a substrate for SCOT, and Leu, as a major activator of anabolic signaling, to compare anabolic and catabolic signals. The expression of genes related to the ubiquitin-proteasome system, which peaked in fasted cells after 5 h of starvation, was suppressed upon treatment with Glc, Gln, or αKG (for Atg1), and with Gln, LA, or αKG (for Murf-1) (Fig. 6A). The increase in AMPK activity after 24 h of starvation was significantly inhibited by the administration of Glc, Gln, LA, or αKG, but not by that of βOHB or Leu (Fig. 6B). After 24 h of starvation, the enhancement of autophagy, as assessed by the LC3II/LC3I ratio, was inhibited by Glc, LA, or αKG treatment, while the inhibition of protein synthesis, as assessed by phosphorylation of p70S6K, was restored by treatment with all the tested substrates, except βOHB (Fig. 6B).

Additionally, we evaluated the effect of nutrients, which restored the starvation-induced inhibition of p70S6K phosphorylation on myotube morphology (Fig. 6C). Generally, atrophy occurs when protein breakdown exceeds protein synthesis in muscle cells. Consistent with the results for anabolic and catabolic signaling in the previous section, the incubation of cells in the starvation medium for 24 h led to severe myotube atrophy. Notably, treatment with Glc not only prevented fasting-induced atrophy but also promoted hypertrophic changes in myotubes. Treatment with Gln, LA, or αKG also prevented myotube atrophy but to a lesser extent than that with Glc. Unexpectedly, Leu showed no protective effects against myotube atrophy caused by nutrient starvation.

Finally, we investigated the metabolic pathways through which each nutrient contributes to cellular energy production. In the regular differentiation medium, ATP production in myotubes primarily relied on glycolysis. After 24 h of starvation, glycolytic ATP production significantly decreased, leading to a reduction in total ATP production; however, ATP production by oxidative phosphorylation (OxPhos) increased (Fig. 7). This indicates that nutritional inadequacy triggers a metabolic shift in muscle cells. The starvation-induced decrease in intracellular ATP production was fully restored by the addition of Glc, Gln, LA, or αKG to the starvation medium, resulting in higher intracellular ATP production when compared with that in control cells. This restoration of ATP production was primarily driven by the recovery of glycolysis in Glc-treated cells, whereas it was facilitated by an increase in OxPhos-derived ATP generation in cells treated with Gln, LA, or αKG (Fig. 7).