Figure 1.
COX6A1 and COX6A2 expression in skeletal muscles and heart of WT mice.
Disruption of the gene leads to reduced complex IV activity and enhanced ROS generation. (A) Representative images of western blots for COX6A1 and COX6A2 in different skeletal muscles. COX6A1 expression was only observed in the heart, whereas COX6A2 protein expression is highest in the heart and diaphragm and lowest in the gastrocnemius muscle, suggesting that COX6A2 expression correlates with muscle oxidative capacity. Sol: soleus muscle; Gast: gastrocnemius muscle; Diaph: diaphragm, (B) Complex IV activity measurements in skeletal muscles and heart of WT and Cox6a2−/− mice (n = 3), (C) Oxygen consumption rate in diaphragm of WT vs Cox6a2−/− mice (n = 3). Inhibition of complex IV by NaN3 was used to validate the method, (D) Measurement of steady state ROS levels by DHE staining. Representative images of DHE stained sections of the diaphragm (upper) and hindlimbs (lower) of WT and Cox6a2−/− mice. (E) Quantification of the number of positive myocytes stained with DHE (n = 6). *p<0.05, **p<0.01. Data represent mean+SEM.
Figure 2.
Cox6a2−/− mice are protected against high-fat diet-induced obesity.
(A) Mice were fed a regular diet or a HFD starting from 6 weeks of age (n = 5 for WT, n = 14–16 for Cox6a2−/− mice). Body weight was monitored weekly. The right panel shows representative images of WT and Cox6a2−/− mice after 12 weeks of HFD feeding, (B) Absolute food intake, relative food intake and feed efficiency of mice fed a HFD (n = 5), (C) Average weight of gastrocnemius muscle, heart, liver, subcutaneous WAT and gonadal WAT dissected from mice that were fed a HFD for 12 weeks (n = 5). (D) Representative pictures of H&E staining of gonadal WAT of mice fed a HFD for 12 weeks. (E) Quantification of subcutaneous and gonadal WAT cell size and density (n = 5). *p<0.05, **p<0.01, ***p<0.001. In all panels, data represent mean+SEM.
Figure 3.
Increased glucose tolerance and insulin sensitivity in Cox6a2−/− mice is associated with constitutive activation of AMPK.
(A) After a 16–18 h fast, mice were injected with 2.5 mg/g BW glucose and blood glucose levels were monitored for 2 h (n = 7–9 for RD, n = 4–11 for 9 weeks of HFD, n = 3–4 for 22 weeks of HFD). Note that WT animals become progressively glucose intolerant when receiving HFD, whereas Cox6a2−/− mice are completely protected against HFD-induced glucose intolerance, (B) Hyperinsulinemic, euglycemic clamps were performed on mice (21 weeks old) fed a HFD for 15 weeks (n = 6–8). Plasma insulin levels before (10 min: B1; 0 min: B2) and after (70 min: H1; 80 min: H2; 90 min: H3) insulin infusion, (C) The glucose infusion rate (GIR) was monitored for 90 min after administration of a hyperinsulinemic solution via the tail vein, (D) Blood glucose levels before insulin infusion (basal) and at the end of the clamp (hyperinsulinaemia), (E) Whole body glucose disposal (left) and hepatic glucose production (right) were measured during the basal period and under hyperinsulinemic conditions, (F) Western blot analysis of insulin-stimulated phosphorylation of Akt in soleus muscle of regular diet fed mice. No difference was observed between fasted (18 h) WT and Cox6a2−/− mice (n = 3), (G) Western blot analysis of AMPK phosphorylation in response to fasting. Mice (n = 3) on a regular diet were either fed ad libitum or fasted overnight (18 h) before dissection of soleus muscles. *p<0.05, **p<0.01, ***p<0.001. In all panels, data represent mean ± SEM.
Figure 4.
Increased energy expenditure and adaptive thermogenesis in Cox6a2−/− mice.
(A–D) Oxygen consumption (A, left panel), average oxygen consumption (A, right panel), heat generation (B), spontaneous activity (C), and respiratory exchange ratio (D, left panel), average respiratory exchange ratio (D, right panel), measured in weight-matched mice of 12–16 weeks old (n = 4–5). Measurements were made over a three day period. Gray shades in A and B indicate dark cycles, (E–F) Mice (12–14 weeks old) fed a regular diet (E) or a HFD (F) were deprived of food, exposed to cold (4°C) and core body temperature was measured every hour for three hours (n = 5–9 for regular diet, n = 5–11 for HFD), (G) Core body temperature of mice (12–14 weeks old) fed a regular or a HFD (n = 5–9 for regular diet, n = 5–11 for HFD). *p<0.05, **p<0.01, ***p<0.001. In all panels, data represent mean ± SEM.
Figure 5.
Increased UCP2 (A) and UCP1 (B) expression in metabolically active tissues of Cox6a2−/− mice.
Quantitative RT-PCRs were performed on cDNA from gastrocnemius muscle, diaphragm, soleus muscle, heart, white and brown adipose tissue (n = 3–5). Gene expression in WT mice was set at 1.0 for each individual tissue. UD = Undetectable. Please note that UCP1 mRNA expression in BAT is about 100-fold that of other tissues. *p<0.05, **p<0.01, ***p<0.001. In all panels, data represent mean ± SEM.
Figure 6.
Fiber type switch and increased mitochondrial size in muscles of Cox6a2−/− mice.
(A) Heat maps of top subsets of genes of a Gene Set Enrichment Analysis of gene expression in diaphragm of WT versus Cox6a2−/− mice on a regular diet (n = 3). Genes were ranked according to their signal-to-noise ratio. Left: Electron transport chain genes. Right: Striated muscle contraction genes. Red color indicates high expression, green color indicates low expression. KO: Cox6a2−/−, (B) Relative changes in expression of genes of the electron transport chain in diaphragm and gastrocnemius muscle of Cox6a2−/− mice compared to WT mice. Horizontal bars indicate mean difference in expression. For complex IV, Cox6a2 was not taken into account for calculation of the mean difference in expression. CI, complex I; CII, complex II; CIII, complex III; CIV, complex IV; CV, complex V (ATP synthase); CytC, cytochrome c, (C) Fiber typing of gastrocnemius muscle by a metachromatic ATPase assay. ATPase activity stains type I fibers dark blue, type IIA fibers light blue, and type IIB fibers very light blue. Representative images of fiber typing in the gastrocnemius muscle from WT and Cox6a2−/− mice are shown. Scale bar = 100 µm. I, IIA, and IIB refer to type I, type IIA, and type IIB fibers, (D) Quantification of muscle fiber types in gastrocnemius muscle of WT and Cox6a2−/− mice (n = 20–30). Note the increase in the total number of fibers in gastrocnemius muscle from Cox6a2−/− mice, (E) Cross-sectional area of different fiber types in gastrocnemius of WT and Cox6a2−/− mice (n = 30–60), (F–G) Transmission electron micrographs of WT and Cox6a2−/− diaphragm. Perinuclear (F) and intermyofibrillar mitochondria (G) are shown. N: nucleus; M: mitochondrion; IM: intermyofibrillar mitochondria; Z: small intermyofibrillar mitochondria located near Z-discs, (H) Quantification of the mitochondrial size of the diaphragm of WT and Cox6a2−/− mice, (I) Pgc-1α mRNA expression in diaphragm of WT and Cox6a2−/− mice (n = 3) was measured using qRT-PCR, (J) mRNA expression signals for Mfn1 and Opa1 as measured by microarrays.
Figure 7.
Isolated soleus muscle from Cox6a2−/− mice is more resistant to fatigue.
(A) Muscle tension measurements. Twitch (both 1 Hz) and tetanic (50 Hz and 100 Hz) tension were measured from isolated soleus and EDL muscle, respectively (n = 8–10), (B) Grip strength measurements (n = 4–5), (C) 14 week old WT and Cox6a2−/− mice were forced to run up- or downhill until exhaustion. Data are presented as distance (m) ran until exhaustion (n = 5), (D) Muscle force during a fatigue test and after recovery in isolated soleus muscle (n = 8–10), (E) Skeletal muscle ATP content. ATP concentrations were measured in diaphragm, gastrocnemius and soleus muscle of WT and Cox6a2−/− mice (n = 5). *p<0.05, **p<0.01, ***p<0.001. In all panels, data represent mean ± SEM.
Figure 8.
Model explaining the metabolic phenotype of Cox6a2−/− mice.
Loss of COX6A2 protein (ΔCOX6A2) in complex IV enhances ROS production by the respiratory chain, which activates AMPK, PGC-1α and upregulates uncoupling protein expression in skeletal muscle. This results in increased energy expenditure, non-shivering thermogenesis, as well as muscle fiber type switch and enhanced insulin sensitivity.