Cardiomyocyte-specific deficiency of ketone body metabolism promotes accelerated pathological remodeling

Abstract

Objective: Exploitation of protective metabolic pathways within injured myocardium still remains an unclarified therapeutic target in heart disease. Moreover, while the roles of altered fatty acid and glucose metabolism in the failing heart have been explored, the influence of highly dynamic and nutritionally modifiable ketone body metabolism in the regulation of myocardial substrate utilization, mitochondrial bioenergetics, reactive oxygen species (ROS) generation, and hemodynamic response to injury remains undefined.

Methods: Here we use mice that lack the enzyme required for terminal oxidation of ketone bodies, succinyl-CoA:3-oxoacid CoA transferase (SCOT) to determine the role of ketone body oxidation in the myocardial injury response. Tracer delivery in ex vivo perfused hearts coupled to NMR spectroscopy, in vivo high-resolution echocardiographic quantification of cardiac hemodynamics in nutritionally and surgically modified mice, and cellular and molecular measurements of energetic and oxidative stress responses are performed.

Results: While germline SCOT-knockout (KO) mice die in the early postnatal period, adult mice with cardiomyocyte-specific loss of SCOT (SCOT-Heart-KO) remarkably exhibit no overt metabolic abnormalities, and no differences in left ventricular mass or impairments of systolic function during periods of ketosis, including fasting and adherence to a ketogenic diet. Myocardial fatty acid oxidation is increased when ketones are delivered but cannot be oxidized. To determine the role of ketone body oxidation in the remodeling ventricle, we induced pressure overload injury by performing transverse aortic constriction (TAC) surgery in SCOT-Heart-KO and αMHC-Cre control mice. While TAC increased left ventricular mass equally in both groups, at four weeks post-TAC, myocardial ROS abundance was increased in myocardium of SCOT-Heart-KO mice, and mitochondria and myofilaments were ultrastructurally disordered. Eight weeks post-TAC, left ventricular volume was markedly increased and ejection fraction was decreased in SCOT-Heart-KO mice, while these parameters remained normal in hearts of control animals.

Conclusions: These studies demonstrate the ability of myocardial ketone metabolism to coordinate the myocardial response to pressure overload, and suggest that the oxidation of ketone bodies may be an important contributor to free radical homeostasis and hemodynamic preservation in the injured heart.

Keywords: Mitochondrial metabolism; Myocardial ketone body metabolism; Nuclear magnetic resonance (NMR) measurement of substrate metabolism; Oxidative stress; Ventricular remodeling.

Figure 1

Myocardial SCOT expression is required for ketone body oxidation. (A) Immunoblot for ketolytic enzyme succinyl-CoA:3-oxoacid-CoA transferase (SCOT) and actin in heart and skeletal muscle protein lysates derived from adult αMHC-Cre, Oxct1^flox/flox, and SCOT-Heart-KO mice. Immunoblot (B) and quantification (C) for ketolytic enzyme succinyl-CoA:3-oxoacid-CoA transferase (SCOT) and actin in heart protein lysates derived from adult αMHC-MerCreMer and tamoxifen-inducible SCOT-Heart-KO mice following intraperitoneal injections of 20 mg/kg tamoxifen for 21 d. (D) ¹³C-edited proton NMR spectrum (2.1–2.5 ppm, relative to chemical shift of trimethylsilyl propionate internal standard) from hearts of αMHC-Cre and SCOT-Heart-KO mice that had been perfused with a buffer containing 11 mM glucose, 20 μU/mL insulin, 0.6 mM palmitate, and 1 mM sodium d-[2,4-¹³C₂]βOHB for 15 min prior to collection of tissues and generation of extracts. ¹³C-Glutamate is a reporter of the contribution of ¹³C-labeled d-βOHB to TCA cycle flux, and is selectively absent in myocardial extracts from SCOT-Heart-KO mice. (E) Fractional ¹³C-enrichment of glutamate (a surrogate for the contribution of a ¹³C-labeled substrate to the TCA cycle) in hearts of αMHC-Cre and SCOT-Heart-KO mice perfused ex vivo with a buffer containing either (i) 1.2 mM [U-¹³C]palmitic acid, 5 mM glucose, 1 μU/mL insulin, 1 mM lactate and 0 mM or 1 mM unlabeled d-βOHB, or (ii) 0.6 mM [U-¹³C]palmitic acid, 11 mM glucose, 10 μU/mL insulin, 1 mM lactate and 0.25 mM unlabeled d-βOHB measured by ¹³C-edited proton NMR. n = 4–9 mice/group. **p ≤ 0.01, ***p ≤ 0.001 by 1-way ANOVA with Tukey's post hoc analysis.

Figure 2

Reconfigured metabolism in the absence of myocardial SCOT. (A, B) Fractional % ¹³C-enrichment of glutamate (glu), lactate (lac), and alanine (ala) or absolute pools of metabolites (nmol/mg tissue) in hearts of αMHC-Cre control and SCOT-Heart-KO mice perfused ex vivo with a buffer containing 11 mM [1-¹³C]glucose, 20 μU/mL insulin, 0.6 mM palmitate, 0.25 mM d-βOHB and 1 mM sodium lactate measured by ¹³C-edited proton NMR. (C) Myocardial glucose uptake measured by perfusing hearts with a buffer containing 1 mM [6-¹³C]-deoxyglucose, 20 μU/mL insulin, 0.6 mM palmitate and 1 mM sodium lactate and the phosphorylation of 2-deoxy-[6-¹³C]glucose was detected by ¹³C-edited proton NMR. (D) Myocardial glycogen content. (E, F) Fractional ¹³C-enrichment of glutamate, lactate, and alanine (%), or total pools of metabolites (nmol/mg tissue) in hearts following perfusion with 11 mM [1-¹³C] glucose, 20 μU/mL insulin, 0.6 mM palmitate and 0.25 mM d-βOHB, and 0 mM lactate; or (G, H) 11 mM glucose, 20 μU/mL insulin, 0.6 mM palmitate, 0.25 mM d-βOHB and 1 mM sodium [3-¹³C]-lactate measured by ¹³C-edited proton NMR. (I) A surrogate for anaplerosis was acquired by perfusing hearts with a buffer containing 11 mM glucose, 10 μU/mL insulin, 0.6 mM palmitate, 0.25 mM d-βOHB, 1 mM lactate and 0.25 mM sodium [1-¹³C]-acetate. n = 5–10 mice/group. *p ≤ 0.05, ***p ≤ 0.001 by 1-way ANOVA with Tukey's post hoc test.

Figure 3

SCOT deficiency is associated with worsened pathological remodeling following pressure overload surgery. SCOT-Heart-KO mice and αMHC-Cre controls underwent transverse aortic constriction (TAC) surgery to promote pressure overload-induced pathological remodeling and were analyzed following either 4 wk or 8 wk TAC. (A) Left ventricular mass (LVM) index, mg LV/mm tibia length; (B) Mean pressure gradient across the aortic arch, mmHg; (C) LV relative wall thickness (RWT); and (D) end diastolic volume (EDV) index, mL/mm tibia length, and ejection fraction, %, were all assessed by echocardiography. n = 5–10 mice/group. *p ≤ 0.05, **p ≤ 0.01 compared to αMHC-Cre mice at same time point; ^#p ≤ 0.05, ^##p ≤ 0.01, ^###p ≤ 0.001, ^####p ≤ 0.0001 compared to baseline time point by 2-way ANOVA with Bonferroni post hoc analysis (E) Representative H&E stained heart cross sections and higher power images. Scale bars are 50 μM.

Figure 4

Pressure overload-induced ultrastructural abnormalities SCOT-Heart-KO mice. (A, B, D, E) Lower power transmission electron microgrpahs reveal normal cardiac muscle structure in sham operated SCOT-Heart-KO mice. (C, F) High power images show no differences in mitochondrial ultrastructure between genotypes. (G, H, J, K) Following 4 wk TAC, the structural integrity of the myocardium appears compromised in SCOT-Heart-KO mice with evident breakdown of the myofibrils as well as thickening of the Z-lines. (I, L) Despite myofibrilar disarray, mitochondrial ultrastructure appears to be preserved in SCOT-Heart-KO mice. Scale bars are 2 μm (panels A, D, G, J), 1 μM (B, E, H, K), and 200 nm (C, F, I, L). n = 2 mice/group.

Figure 5

Relative genome content and respiration studies of cardiac mitochondria. (A, B) Quantification of mitochondrial genome copy number (relative abundance) by qPCR using purified heart gDNA from SCOT-Heart-KO and αMHC-Cre mice at baseline, or after 4 wk or 8 wk TAC. Data are presented as means ± SEM; n = 4–5/group, *p ≤ 0.05 by 1-way ANOVA with Tukey's post hoc analysis. (C) Increased myocardial expression of mitochondrial transcription factor A (Tfam). (D, E) Respiration rates in the basal leak condition (state 2), ADP-stimulated condition (state 3), F₁F₀-ATPase independent condition (state 4, oligomycin), and uncoupled condition (FCCP) in mitochondria isolated from left ventricles of SCOT-Heart-KO and αMHC-Cre mice using palmitoyl-l-carnitine and malate or succinate and rotenone as substrates at (D) baseline or (E) following 4 wk TAC. Data are presented as means ± SEM; n = 4–7 mice/group, *p ≤ 0.05, ***p ≤ 0.001 by 2-way ANOVA with Bonferroni post hoc analysis. (F, G) No differences in NAD⁺, NADH, NADt or NAD⁺/NADH ratios were detected when comparing whole myocardial lysates of SCOT-Heart-KO and αMHC-Cre control mice at (F) baseline or (G) after 4 wk TAC. n = 5–7 mice/group.

Figure 6

Decreased acetylation of lysine residues in myocardial proteins of SCOT-Heart-KO mice. Decreased acetylation of lysine residues is evident at baseline in both (A) whole myocardial and (B) isolated mitochondrial protein extracts as well as (C) after 4 wk TAC in whole myocardial extracts. n = 5–7 mice/group, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001 by Student's t test.

Figure 7

Pressure-overload-induced signatures of increased oxidative stress in SCOT-Heart-KO mice. (A) H₂O₂ emission was measured in isolated mitochondria following the addition of metabolic substrate (i) succinate, (ii) succinate plus d-βOHB, or (iii) succinate plus acetoacetate (AcAc); rotenone was added to eliminate further H₂O₂ emission. *p = 0.02 and p = 0.04 for rates of H₂O₂ emission following the addition of succinate plus AcAc in SCOT-Heart-KO and αMHC-Cre mitochondria, respectively, when compared with the rates of H₂O₂ emission following the addition of succinate alone, n = 7 mice/group. (B) Tissue superoxide content was quantified in situ using fluorogenic probes CellROX^® Green, detecting mitochondrial and nuclear ROS and CellROX^® Deep Red reagent detecting cytosolic ROS. Relative intensity was quantified using ImageJ. Sham operated images not shown. (C) Protein oxidation was quantified in whole myocardial lysates of SCOT-Heart-KO and αMHC-Cre mice via Western blotting following a reaction by 2,4-dinitrophenylhydrazine (DNPH). Sham operated blot not shown. n = 3–7 mice/group. ***p ≤ 0.001, ****p ≤ 0.0001 by 2-way ANOVA with Bonferroni post hoc analysis.