Model of Acot2 depletion in muscle and study design

A. Proteomics analysis from single murine muscle fibers. One ANOVA and multiple comparisons with correction for FDR (0.05; Benjamini, Krieger, Yekutieli method): * 0.05 ≤ q ≤ 0.03, ** 0.01 ≤ q < 0.03, *** q<0.01.

B. Representative immunoblots comparing Acot2 levels in skeletal muscle (SM) mitochondria from FL/FL (Loxp/Loxp control) and Acot2-/- (knockdown of Acot2 in striated muscle) mice. SDHA (succinate dehydrogenase) was used as the loading control.

C. Thioesterase activity measured in 40 µg of FL/FL and Acot2-/- SM mitochondria per reaction with 10 µM malonyl-CoA as substrate in n=10/genotype.** p=0.02, unpaired t-test.

D. Summary of the experimental design.

E. Short-, medium-, long-chain acyl CoAs and CoASH detected by high resolution mass spectrometry analysis of quadriceps muscle from FL/FL and Acot2-/- mice studied under chow, fasting, running after fasting, and HFD conditions, n=6/genotype. Statistics: two-way ANOVA and multiple compariFsons (q values) with correction for FDR (0.05; Benjamini, Krieger, Yekutieli method): * 0.05 ≤ q ≤ 0.04, ** 0.01 ≤ q < 0.03, *** q<0.01. N.S.: not significant.

F. Ratio of CoASH to the sum of ACoAs (data from panel E). Statistics as in Panel E.

Panels A, C, E, F: Bars are mean ± s.e.m, individual points are data from each mouse.

Lack of change, in any condition, in RNA or protein abundance of components of major CoA-related metabolic pathways in Quad from Acot2-/- mice

A-C. RNAseq analysis of Quad muscle, from mice fed normal chow, overnight fasted, or fed HFD for 7 days. Shown are transcript levels for enzymes in acyl-CoA metabolism (A), CoA synthesis/ degradation pathways (B), and fatty acid binding proteins (C), in Quad from chow-fed, overnight fasted and high fat diet (HFD)-fed mice. Each box is the average of n=4/genotype.

D. Representative immunoblots from Quad lysates from chow-fed, fasted and HFD-fed mice, showing expression of VLCAD, β-HAD, ETFa, with PDH, GRB2 as loading controls; as well as CoASY and Nudt8, with PHD as loading control. At right: bar charts showing quantification. Values: mean ± s.e.m., points show individual mice; n=4/genotype/condition. N.S. Not significant (unpaired t-test).

β-oxidation in mitochondria from white SM is stimulated or inhibited by Acot2 loss, depending on lipid supply

A. Oxygen consumption (JO2) in mitochondria from white SM, isolated from chow-fed, and supplied with 20 µM PCarn and 0.1 mM malate. The creatine kinase clamp approach was used to simulate changes in ATP demand (ADP/ATP) by varying phosphocreatine (PCr) concentration. Values: mean ± s.e.m..

B. The data from (A) were expressed as a function of Gibbs free energy for ATP (ΔGATP). The slope reflects the conductance. Right panel: conductance; bar chart are mean ± s.e.m., individual points are from individual mice; p value: unpaired t-test, n=7/genotype.

C. Left panel: JO2 measured in white SM mitochondria, with saturating [ADP] and 20 or 40 µM PCarn (plus 0.1 mM malate). Statistics: two-way ANOVA and multiple comparisons (q values) with correction for FDR (0.05; Benjamini, Krieger, Yekutieli method): * 0.05 ≤ q ≤ 0.03, ** 0.01 ≤ q < 0.03; n=7/genotype. Right panel: bar chart showing the difference in JO2 (ΔJO2) for 40 vs. 20 µM PCarn. p value: unpaired t-test, n=7/genotype.

D. Left panel: JO2 measured in white SM mitochondria, with saturating [ADP] and 20 µM PCarn (plus 0.1 mM malate), showing measurements 6 and 18 minutes after ADP addition. Note: 6 minutes data for both genotypes are the same data shown in Panel C (20 µM PCarn) Statistics: two-way ANOVA and multiple comparisons (q values) with correction for FDR (0.05; Benjamini, Krieger, Yekutieli method): *** q < 0.01; n=7/genotype. Right panel: bar chart showing the difference in JO2 (ΔJO2) for 18 vs. 6 minutes; p value: unpaired t-test, n=7/genotype.

E. Left panel: JO2 measured in white SM mitochondria, with saturating [ADP], 10 mM pyruvate and 5 mM malate (saturating [substrate]), measured 6 minutes after ADP addition. Middle panel: Difference in ADP-driven JO2 (ΔJO2) for 18 vs. 6 minutes Right panel: JO2 after addition of oligomycin to inhibit the ATP synthase. All panels: p value: unpaired t-test, n=7/genotype.

F. JO2 measured in white SM mitochondria fueled by 20 or 40 µM PCarn (and 0.1 mM malate), after addition of oligomycin to inhibit the ATP synthase. Statistics: two-way ANOVA and multiple comparisons (q values) with correction for FDR (0.05; Benjamini, Krieger, Yekutieli method): ** 0.01 ≤ q < 0.03; n=7/genotype.

G. Left panel: scheme illustrating the approach to evaluate proton leak in isolated mitochondria. Mitochondrial membrane potential (Δψm) and JO2 in the presence of oligomycin were measured in parallel in each preparation of mitochondria from white SM. Antimycin (inhibits Complex III of the ETC) was used to decrease the driving force. Middle panel: Proton leak measured in mitochondria fueled with 20 µM PCarn and 0.1 mM malate. Right panel: Same experiment as in Middle panel, except using 40 µM PCarn and 0.1 mM malate. Middle and right panels: insets show bar chart of mean (±s.e.m.) Δψm and JO2 at the arrow; p value: unpaired t-test, n=6/genotype.

Panels B, C, D, E, F: Bar chart: mean ± s.e.m., individual points show data from each mice.

β-oxidation intermediates provide evidence for overload in white SM from Chow-fed and fasted Acot2-/- mice

A. Schematic representation of β-oxidation pathway and Acot2 reaction.

B. Relative abundance of the β-oxidation intermediates for each reaction (1-4, corresponding to steps 1-4 shown in (A)) for all chain lengths (indicated by C4, C6 etc). Metabolites were measured by high resolution mass spectrometry. Statistics: shown within the panel; BKY: Benjamini, Krieger, Yekutieli, to correct for multiple comparisons, FDR: p=0.05. n=6/genotype/condition.

C. Pearson’s correlation coefficient (r) analysis between adjacent steps of β-oxidation, using the intermediates shown in (B). Roman numerals: enzymes for each step (see panel A).

D. Analysis of Pearson’s correlation coefficients. Statistics: shown within the panel; BKY: Benjamini, Krieger, Yekutieli, to correct for multiple comparisons, FDR: p=0.05.

Indirect calorimetry shows a switch to greater glucose oxidation in Acot2-/- mice

A. Time course (time of the day) showing whole-body O2 consumption (VO2), CO2 production (VCO2) and respiratory quotient (RQ), during the light and dark phase of the day, in chow-fed mice. Bar chart shows VO2, VCO2 and RQ values in different activity bins (1 is lowest activity bin, 3 is the highest). For each mouse, data were averaged over 2 days, after a day of acclimation.

B. Same as (A), but in overnight-fasted mice; food was removed at 7pm (19-h).

C. Same as (A), but in mice fed 60% HFD for 7 days; data reflect the last 24 hrs of HFD.

In all panels: Time course measurements show mean ± s.e.m. Bar charts show mean ± s.e.m., and individual points show data from each mice. Statistics: two-way ANOVA, Statistics: two-way ANOVA and multiple comparisons (q values) with correction for FDR (0.05; Benjamini, Krieger, Yekutieli method): * 0.05 ≤ q ≤ 0.03, ** 0.01 ≤ q < 0.03; n=6/genotype/condition.

Less lipid accumulation in Quad and faster glucose disposal in HFD-fed Acot2-/- mice

A. Abundance of lipids, measured by high-resolution LC-MS analysis, in Quads from chow- and HFD-fed mice, n=6/genotype. Statistics are shown in the panel. BKY: Benjamini, Krieger, Yekutieli, to correct for multiple comparisons, FDR: p=0.05; * 0.05 ≤ q ≤ 0.03, ** 0.01 ≤ q < 0.03, *** q<0.01.

B. Triacylglycerols measured in liver, heart and serum, measured biochemically, in chow and HFD fed mice, n=7-11/genotype. Statistics: two-way ANOVA and multiple comparisons (q values) with correction for FDR (0.05; Benjamini, Krieger, Yekutieli method): ** 0.01 ≤ q < 0.03.

C. Oral glucose tolerance test (GTT, 2 µg/ 1 g of body weight glucose injection) in chow-fed mice, n=11 FL/FL and n=10 Acot2-/-. N.S.: not significant, unpaired t-test.

D. Serum insulin levels 15 min after oral glucose (2 µg/ 1 g of body weight) injection in chow-fed mice, n=7 FL/FL and n=8 Acot2-/-. N.S.: not significant, unpaired t-test.

E. Insulin tolerance test (ITT, 1.5 U/ kg of body weight insulin injection) in chow-fed mice, n=4/genotype. N.S.: not significant, unpaired t-test.

F. Oral glucose tolerance test (GTT, 1 µg/ 1 g of body weight glucose injection) in HFD-fed mice, n=11/genotype. P value: unpaired t-test.

G. Serum insulin levels before (in n=9/genotype) and 15min after (n=8 FL/FL, n=10 Acot2-/-) oral glucose (1 µg/ 1 g of body weight) injection in HFD-fed mice. Statistics: two-way ANOVA and multiple comparisons (q values) with correction for FDR (0.05; Benjamini, Krieger, Yekutieli method).

H. Insulin tolerance test (ITT, 3 U/ kg of body weight insulin injection) in HFD-fed mice, n=9 FL/FL and n=8 Acot2-/-. P value: unpaired t-test.

Panel A-F: values are mean ± s.e.m. The points represent individual experiments.