Loss of mitochondrial calcium uniporter rewires skeletal muscle metabolism and substrate preference

Abstract

Skeletal muscle mitochondria readily accumulate Ca²⁺ in response to SR store-releasing stimuli thanks to the activity of the mitochondrial calcium uniporter (MCU), the highly selective channel responsible for mitochondrial Ca²⁺ uptake. MCU positively regulates myofiber size in physiological conditions and counteracts pathological loss of muscle mass. Here we show that skeletal muscle-specific MCU deletion inhibits myofiber mitochondrial Ca²⁺ uptake, impairs muscle force and exercise performance, and determines a slow to fast switch in MHC expression. Mitochondrial Ca²⁺ uptake is required for effective glucose oxidation, as demonstrated by the fact that in muscle-specific MCU^-/- myofibers oxidative metabolism is impaired and glycolysis rate is increased. Although defective, mitochondrial activity is partially sustained by increased fatty acid (FA) oxidation. In MCU^-/- myofibers, PDP2 overexpression drastically reduces FA dependency, demonstrating that decreased PDH activity is the main trigger of the metabolic rewiring of MCU^-/- muscles. Accordingly, PDK4 overexpression in MCU^fl/fl myofibers is sufficient to increase FA-dependent respiration. Finally, as a result of the muscle-specific MCU deletion, a systemic catabolic response impinging on both liver and adipose tissue metabolism occurs.

Fig. 1

Constitutive muscle-specific MCU deletion decreases muscle performance and causes fiber-type switching. a Mlc1f-Cre mice were crossed with mice in which exon 5 of the MCU gene is flanked by loxP sequences. Expression of the Cre recombinase in embryogenesis determines the deletion of the MCU gene specifically in skeletal muscle. For the sake of simplicity, throughout the manuscript the Mlc1f-Cre::MCU^fl/fl mice are named skMCU^−/−. b Western blot analyses demonstrated efficient MCU deletion in hindlimb muscles. c Cytosolic and mitochondria [Ca²⁺] were simultaneously measured in adult FDB myofibers. While in MCU^fl/fl myofibers caffeine-induced cytosolic Ca²⁺ increase was followed by mitochondrial Ca²⁺ uptake (left representative traces), in skMCU^−/− mice mitochondrial Ca²⁺ uptake was abolished (right representative traces). Histogram on the right reports mean mitochondrial Ca²⁺ peaks upon caffeine treatment. *p < 0.05, t test (two-tailed, unpaired) of >20 fibers per condition. Data are presented as mean ± SD. d Ratiometric measurements of cytosolic Ca²⁺ transients upon caffeine treatment highlighted a reduction in peak cytosolic [Ca²⁺] of skMCU^−/− FDB myofibers compared to MCU^fl/fl controls. *p < 0.05, t test (two-tailed, unpaired) of >20 fibers per condition. Data are presented as mean ± SD. Representative traces are reported on the right side. e Resting cytosolic [Ca²⁺] was unaltered in skMCU^−/− FDB myofibers compared to MCU^fl/fl controls. Data are presented as mean ± SD (>20 fibers per condition). f Fiber size was reduced in skMCU^−/− hindlimb muscles compared to controls. *p < 0.05, t test (two-tailed, unpaired) of four animals per condition. Data are presented as mean ± SD. g Mean maximal running distance in a single bout of run on a treadmill of MCU^fl/fl and skMCU^−/− mice indicated that MCU deletion negatively affects exercise performance. *p < 0.05, t test (two-tailed, unpaired) of four animals per condition. Data are presented as mean ± SD. h Immunofluorescence analyses of soleus myofibers with antibodies specific for different myosin heavy chains (anti-MyHC1 and anti-MyHC2A) demonstrated a reduction in slow-twitch type 1 myosin-positive fibers and an increase in fast-twitch type 2A myosin-positive fibers in skMCU^−/− samples. *p < 0.05, Mann–Whitney rank-sum test (two-tailed, unpaired) of four animals per condition. Data are presented as mean ± SD. Representative images are reported on the right side. Scale bar 100 μm. i In vivo fatigue of gastrocnemius muscles was measured in skMCU^−/− mice compared to MCUfl/fl mice. Six animals per condition. Data are presented as mean ± SD (five animals per condition). Mean maximal running distances after each of 3 days of downhill run on a treadmill of MCUfl/fl and skMCU^−/− mice. *p < 0.05, t test (two-tailed, unpaired) of five animals per condition. Data are presented as mean ± SD

Fig. 2

Muscle-specific MCU deletion triggers a metabolic shift toward fatty acid utilization. a OCR measurements indicated reduced respiratory capacity in skMCU^−/− FDB myofibers compared to controls. Left: representative traces. Right: quantification. To calculate basal and maximal respiration, non-mitochondrial O₂ consumption was subtracted from absolute values. ATP linked respiration was calculated as the difference between basal and oligomycin-insensitive O₂ consumption. Data are normalized on mean Calcein fluorescence. *p < 0.05, t test (two-tailed, unpaired) of ten samples per condition. Data are presented as mean ± SD. b Blood lactate concentration was increased in skMCU^−/− mice compared to controls. *p < 0.05, t test (two-tailed, unpaired) of four animals per condition. Data are presented as mean ± SD. c The rate of glycolysis was determined by measuring the conversion of 5-³H-glucose to ³H₂O. skMCU^−/− TA myofibers showed increased glycolysis rate relative to controls. *p < 0.05, t test (two-tailed, unpaired) of four animals per condition. d Phosphorylation levels of PDH were increased in skMCU^−/− TA muscles compared to controls. e Glucose dependent-respiration was decreased in skMCU^−/− FDB myofibers compared to controls. Left: OCR measurements were performed as in a. UK5099 was added to inhibit the mitochondrial pyruvate carrier (MPC). Right: Glucose dependency was calculated and expressed as the percentage of basal OCR. Data are normalized on mean Calcein fluorescence. *p < 0.05, t test (two-tailed, unpaired) of five samples per condition. Data are presented as mean ± SD. f Fatty acid (FA) content was increased in skMCU^−/− TA muscles relative to controls. *p < 0.05, t test (two-tailed, unpaired) of six animals per condition. Data are presented as mean ± SD. g FA-dependent respiration was increased in skMCU^−/− FDB myofibers compared to controls. Left: OCR measurements were performed as in a. Etomoxir was added to inhibit FA utilization. Right: FA dependency was calculated and expressed as the percentage of basal OCR. Data are normalized on mean Calcein fluorescence. *p < 0.05, t test (two-tailed, unpaired) of five samples per condition. Data are presented as mean ± SD. h OCR measurement demonstrated differential substrates utilization in skMCU^−/− myofibers. Left: during OCR measurements UK5099 was added to freshly isolated myofiber followed by etomoxir. Right: Glucose dependency and FA dependency were calculated and expressed as the percentage of basal OCR. Data are normalized on mean Calcein fluorescence. *p < 0.05, t test (two-tailed, unpaired) of five samples per condition. Data are presented as mean ± SD. i Phosphorylation levels of ACC were increased in skMCU^−/− TA muscles compared to controls. j Untargeted metabolomics analysis of skMCU^fl/fl and skMCU^−/− muscles. Metabolites set enrichment analysis revealed significant upregulation of acylcarnitine and diacylglycerol lipids in skMCU^−/− muscles (pathway p values are reported). Metabolites that are significantly increased are labeled. The size of the circles is proportional to the fold change (fold change values are reported in Table 1). p Value is shown in the figure (Welch’s two-sample t test of nine animals per condition)

Fig. 3

Pyruvate dehydrogenase inhibition underlies the increase in FA oxidation of skMCU^−/− muscles. a Western blot analysis showed that Flag-PDK4 overexpression in MCU^fl/fl FDB muscles increased the phosphorylation levels of PDH (pcDNA3.1 was used as control). b FA-dependent respiration was increased in MCU^fl/fl PDK4-overexpressing FDB myofibers. Left: OCR measurements were performed as in Fig. 2a. Etomoxir was added to inhibit FA utilization. Right: FA dependency was calculated and expressed as the percentage of basal OCR. Data are normalized on mean Calcein fluorescence. *p < 0.05, t test (two-tailed, unpaired) of five samples per condition. Data are presented as mean ± SD. c Western blot analysis showed that Flag-PDP2 overexpression in skMCU^−/− FDB muscles decreased the phosphorylation levels of PDH (pcDNA3.1 was used as control). d FA-dependent respiration was decreased in skMCU^−/− PDP2-overexpressing FDB myofibers. Left: OCR measurements were performed as in Fig. 2a. Etomoxir was added to inhibit FA utilization. Right: FA dependency was calculated and expressed as the percentage of basal OCR. Data are normalized on mean Calcein fluorescence. *p < 0.05, t test (two-tailed, unpaired) of five samples per condition. Data are presented as mean ± SD

Fig. 4

Muscle-specific MCU deletion impinges on systemic metabolism. a Glucose tolerance test of MCU^fl/fl and skMCU^−/− mice. skMCU^−/− mice had decreased glycemia at all time points compared to MCU^fl/fl *p < 0.05, t test (two-tailed, unpaired) of five animals per condition. Data are presented as mean ± SD. b In vivo measurements of 2-deoxy-d-glucose uptake revealed that skMCU^−/− TA and SOL muscles accumulated more glucose than MCU^fl/fl muscles. *p < 0.05, t test (two-tailed, unpaired) of five animals per condition. Data are presented as mean ± SD. c Glycogen content was increased in skMCU^−/− muscles relative to controls. *p < 0.05, Mann–Whitney rank-sum test (two-tailed, unpaired) of four animals per condition. Data are presented as mean ± SD. d In vivo measurements of 2-deoxy-d-glucose uptake revealed that skMCU^−/− liver took up less glucose than controls. *p < 0.05, t test (two-tailed, unpaired) of five animals per condition. Data are presented as mean ± SD. e Relative mRNA expression levels of the gluconeogenesis enzymes G6pase and Pck1 in skMCU^−/− liver compared to MCU^fl/fl liver. *p < 0.05, t test (two-tailed, unpaired) of nine animals per condition. Data are presented as mean ± SD. f Glycogen content was decreased in skMCU^−/− mice liver compared to controls. *p < 0.05, Mann–Whitney rank-sum test (two-tailed, unpaired) of six animals per condition. Data are presented as mean ± SD. g Phosphorylation levels of GSK3 α/β and AKT were decreased in skMCU^−/− livers compared to controls. Total AKT and GSK3 α/β were used as control. h Relative mRNA expression levels of the lipases Lipe, Abhd5, Mgll, and Pnpla2 in skMCU^−/− liver compared to MCU^fl/fl liver. *p < 0.05, t test (two-tailed, unpaired) of nine animals per condition. Data are presented as mean ± SD. i Relative mRNA expression levels of the ketogenesis enzymes Bdh1 and Hmgcs2 in skMCU^−/− liver compared to MCU^fl/fl liver. *p < 0.05, t test (two-tailed, unpaired) of nine animals per condition. Data are presented as mean ± SD. j Plasma beta-hydroxybutyric acid concentration was increased in skMCU^−/− mice compared to controls. *p < 0.05, t test (two-tailed, unpaired) of four animals per condition. Data are presented as mean ± SD. k Plasma FA concentration was increased in skMCU^−/− mice compared to controls. *p < 0.05, t test (two-tailed, unpaired) of four animals per condition. Data are presented as mean ± SD. l The amount of visceral adipose tissue (VAT) was decreased in skMCU^−/− mice compared to MCU^fl/fl animals. *p < 0.05, t test (two-tailed, unpaired) of four animals per condition. Data are presented as mean ± SD. m Relative mRNA expression levels of the lipases Mgll, Lipe, and Pnpla2 in skMCU^−/− VAT compared to MCU^fl/fl liver. *p < 0.05, t test (two-tailed, unpaired) of five animals per condition. Data are presented as mean ± SD

Fig. 5

Inducible muscle-specific MCU deletion impairs skeletal muscle performance and triggers fiber-type switch. a HSA-Cre-ER^T2 mice were crossed with mice in which exon 5 of the MCU gene is flanked by loxP sequences (MCU^fl/fl). At 8 weeks of age, all mice received tamoxifen treatment. Upon tamoxifen binding, Cre recombinase determines the deletion of the MCU gene specifically in the skeletal muscle. For the sake of simplicity, the tamoxifen-treated HSA-Cre-ER^T2::MCU^fl/fl mice are named iskMCU^−/−. b Western blot analyses of different iskMCU^−/− muscles revealed variable downregulation of MCU expression. c Fiber size of iskMCU^−/− hindlimb muscles was reduced proportionally to the degree of MCU ablation detected in b. *p < 0.05, t test (two-tailed, unpaired) of four animals per condition. Data are presented as mean ± SD. d Mean maximal running distance in a single bout of run on a treadmill of MCU^fl/fl and iskMCU^−/− mice indicated that MCU deletion negatively affects exercise performance. *p < 0.05, t test (two-tailed, unpaired) of five animals per condition. Data are presented as mean ± SD. e Immunofluorescence analysis of soleus myofibers with antibodies specific for different myosin heavy chains (anti-MyHC1 and anti-MyHC2A) demonstrated a reduction in slow-twitch type 1 myosin-positive fibers and an increase in fast-twitch type 2A myosin-positive fibers in iskMCU^−/− samples relative to controls. *p < 0.05, Mann–Whitney rank-sum test (two-tailed, unpaired) of four animals per condition. Data are presented as mean ± SD. Representative images are reported on the right side. Scale bar 100 μm

Fig. 6

MCU deletion in adult skeletal muscle impairs whole-body metabolism. a OCR measurements indicated reduced respiratory capacity in iskMCU^−/− FDB myofibers compared to controls. Left: representative traces. Right: quantification. To calculate basal and maximal respiration, non-mitochondrial O₂ consumption was subtracted from absolute values. ATP linked respiration was calculated as the difference between basal and oligomycin-insensitive O₂ consumption. Data are normalized on mean Calcein fluorescence. *p < 0.05, t test (two-tailed, unpaired) of ten samples per condition. Data are presented as mean ± SD. b Phosphorylation levels of PDH were increased in iskMCU^−/− soleus muscles compared to controls. c Glucose dependent-respiration was decreased in iskMCU^−/− FDB myofibers compared to controls. Left: OCR measurements were performed as in a. UK5099 was added to inhibit MPC. Right: Glucose dependency was calculated and expressed as the percentage of basal OCR. Data are normalized on mean Calcein fluorescence. *p < 0.05, t test (two-tailed, unpaired) of five samples per condition. Data are presented as mean ± SD. d FA-dependent respiration was increased in iskMCU^−/−. Left: OCR measurements were performed as in a. Etomoxir was added to inhibit FA utilization. Right: FA dependency was calculated and expressed as the percentage of basal OCR. Data are normalized on mean Calcein fluorescence. *p < 0.05, t test (two-tailed, unpaired) of five samples per condition. Data are presented as mean ± SD. e iskMCU^−/− mice had unaltered fasting glycemia compared to MCU^fl/fl controls. Data are presented as mean ± SD (five animals per condition). f Blood lactate concentration was increased in iskMCU^−/− mice compared to controls. *p < 0.05, t test (two-tailed, unpaired) of four animals per condition. Data are presented as mean ± SD. g Plasma FA concentration was increased in iskMCU^−/− mice compared to controls. *p < 0.05, t test (two-tailed, unpaired) of four animals per condition. Data are presented as mean ± SD. h Plasma beta-hydroxybutyric acid concentration was increased in iskMCU^−/− mice compared to controls. *p < 0.05, t test (two-tailed, unpaired) of four animals per condition. Data are presented as mean ± SD

Fig. 7

Model. a Schematic representation of the metabolic consequences of MCU deletion in the skeletal muscle. Left: Ca²⁺ enters mitochondria through the MCU complex and stimulates the PDH activity. PDH converts pyruvate into acetyl-CoA, which in turn feeds the TCA cycle. Right: in MCU^−/− muscles mitochondrial Ca²⁺ uptake is inhibited, PDH is inactive, and pyruvate oxidation is impaired. In turn, TCA cycle activity decreases. Increased FA metabolism partially compensates the impairment in glucose oxidation. In addition, FA oxidation produces reducing equivalents that fuel the respiratory chain. b The primary skeletal muscle mitochondria defect triggered by MCU deletion has an impact on whole-body metabolism. The decrease in muscle glucose oxidation translates into increased glycolysis, glycogen storage, and lactate production. The aberrant glucose utilization signals to the liver and the adipose tissue, which in turn mobilize nutrient depots. skMCU^−/− mice have decreased muscle force, impaired exercise performance, and display MyHC1 → MyHC2A expression switch