Impaired mitochondrial ketone body oxidation in insulin resistant states

Abstract

Background: Reduced mitochondrial respiratory function has been implicated in metabolic disorders like type 2 diabetes (T2D), obesity, and metabolic dysfunction-associated steatotic liver disease (MASLD), which are tightly linked to insulin resistance and impaired metabolic flexibility. However, the contribution of the ketone bodies (KBs) β-hydroxybutyrate (HBA) and acetoacetate (ACA) as substrates for mitochondrial oxidative phosphorylation (OXPHOS) in these insulin resistant states remains unclear.

Methods: Targeted high-resolution respirometry protocols were applied to detect the differential contribution of HBA and ACA to OXPHOS capacity in heart, skeletal muscle, kidney, and liver of distinct human or murine cohorts with T2D, obesity, and MASLD.

Findings: In humans with T2D, KB-driven mitochondrial OXPHOS capacity was ∼30% lower in the heart (p < 0.05) and skeletal muscle (p < 0.05) compared to non-diabetic controls. The relative contribution of KBs to maximal OXPHOS capacity in T2D was also lower in both the heart (∼25%, p < 0.05) and skeletal muscle (∼50%, p < 0.05). Similarly, in kidney cortex from high-fat diet-induced obese mice, both the absolute and relative contribution of KBs to OXPHOS capacity was ∼15% lower (p < 0.05). Finally, hepatic HBA-driven mitochondrial OXPHOS capacity was 29% lower (p < 0.05) in obese humans with hepatic steatosis compared to humans without.

Interpretation: Mitochondrial KB-driven OXPHOS capacity is impaired in insulin resistant states in various organs in absolute and relative terms, likely reflecting impaired mitochondrial metabolic flexibility. Our data suggest that KB respirometry can provide a sensitive readout of impaired mitochondrial function in diabetes, obesity, and MASLD.

Funding: German Research Foundation, German Diabetes Center, German Federal Ministry of Health, Ministry of Culture and Science of the state of North Rhine-Westphalia, German Federal Ministry of Education and Research, German Center for Diabetes Research, German Heart Foundation, German Diabetes Society, Christiane-and-Claudia Hempel Foundation, European Community and Schmutzler Stiftung.

Keywords: Diabetes mellitus; Ketone bodies; MASLD; Mitochondrial respiration; Obesity.

Fig. 1

Pathways of ketone body-linked ATP generation by mitochondrial oxidative phosphorylation. Ketone body oxidation is initiated by addition of β-hydroxybutyrate (HBA) in the presence of ADP (yellow shaded area); complete ketolysis is achieved by further addition of malate (M; green shaded area); further addition of acetoacetate (ACA) allows further insight into the interplay of the various enzymes involved in ketolysis (for a detailed explanation, see Methods). The isolated contribution of ACA to ketolysis can be achieved by addition of ACA and M in the presence of ADP (blue shaded area). ADP: adenosine diphosphate; ATP: adenosine triphosphate; BDH1: β-hydroxybutyrate dehydrogenase; CoA, coenzyme A; CoQ: coenzyme Q; e⁻: electron; FADH₂: flavin adenine dinucleotide; H⁺: proton; JO₂: oxygen flux; KL: ketolysis; NADH: nicotinamide adenine dinucleotide; _P: phosphorylating (coupled) mitochondrial respiration state; SCOT: Succinyl-CoA: 3-ketoacid (oxoacid) coenzyme A transferase; TCA: tricarboxylic acid cycle. Figure generated using BioRender.com (Toronto, CA).

Fig. 2

Reduced ketone body dependent OXPHOS capacity in the diabetic heart. Mitochondrial respiration (JO₂) of permeabilised ventricular myocardium of heart transplant recipients undergoing routine endomyocardial biopsy without (Control) versus with Type 2 Diabetes (T2D), obtained with (a) the combined ketone body (KB) and (b) the fatty acid oxidation (F), NADH (N), and succinate (S) combined (FNS) pathway mitochondrial respiration protocols. (c) Relative contribution of the different respiratory states of KB-linked versus maximal coupled FNS-linked mitochondrial respiration. (d) [KL]_P/[KL + ACA]_P ratio. a–c: repeated-measurements mixed-effects (REML) model with correction for multiple testing with the false discovery rate. d: Welch's t-test. n = 35 (control) versus 22 (T2D). Data are mean ± SEM. HBA: β-hydroxybutyrate; _E: electron transport chain capacity (noncoupled) mitochondrial respiration state; JO₂: oxygen flux; KL: ketolysis; _L: leak mitochondrial respiration state; _P: phosphorylating (coupled) mitochondrial respiration state.

Fig. 3

Lower KB-dependent mitochondrial OXPHOS capacity in the type 2-diabetic skeletal muscle. Mitochondrial respiration (JO₂) in permeabilised skeletal muscle (vastus lateralis) fibres from type 2 diabetes (T2D) versus glucose-tolerant (Control) individuals, obtained with (a) the combined ketone body and (b) the NADH (N) and succinate (S) combined (NS) pathway mitochondrial respiration protocols. (c) Relative contribution of the different respiratory states of KB-linked versus maximal coupled NS-linked mitochondrial respiration. (d) [KL]_P/[KL + ACA]_P ratio. Analysis of (e) Citrate Synthase (CS) activity, a marker of mitochondrial content, and (f) protein expression of β-hydroxybutyrate dehydrogenase (BDH1) and succinyl-CoA: 3-ketoacid (oxoacid) coenzyme A transferase (SCOT). Representative immunoblots of BDH1 and SCOT from three replicates of each group are shown. a, d: n = 7 versus 13. b, c: n = 6 versus 6, e: n = 6 versus 12, f: n = 7–12. a–c: repeated-measurements mixed-effects (REML) model with correction for multiple testing with the false discovery rate. d and e: Welch's t-test. f: Mann-Whitney-Test. Data are mean ± SEM. HBA: β-hydroxybutyrate; _E: electron transport chain capacity (noncoupled) mitochondrial respiration state; JO₂: oxygen flux; KL: ketolysis; _L: leak mitochondrial respiration state; _P: phosphorylating (coupled) mitochondrial respiration state.

Fig. 4

Decreased KB-dependent OXPHOS capacity in the kidney cortex of diet-induced obesity (DIO) male mice. (a) Mitochondrial respiration (JO₂) of permeabilised kidney cortex from DIO versus normal chow diet-fed (control) C57BL/6J male mice, obtained with the combined ketone body mitochondrial respiration protocol. (b) Relative contribution of the different respiratory states of KB-linked versus maximal coupled mitochondrial respiration determined following subsequent addition of substrates of the NADH (N) and succinate (S) pathways combined ([KL + ACA + NS]_P). (c) [KL]_P/[KL + ACA]_P ratio. (d) Citrate Synthase (CS) activity and (e) protein expression of β-hydroxybutyrate dehydrogenase (BDH1) and succinyl-CoA: 3-ketoacid (oxoacid) coenzyme A transferase (SCOT). (f) Representative immunoblots of BDH1 and SCOT from three control and three DIO mice are shown. n = 10 versus 12. a: repeated-measurements mixed-effects (REML) model with correction for multiple testing with the false discovery rate. b: multiple Mann-Whitney-tests, corrected for multiple testing with the false discovery rate. c–e: Welch's t-test. Data are mean ± SEM. ACA: acetoacetate; HBA: β-hydroxybutyrate; JO₂: oxygen flux; KL: ketolysis; _P: phosphorylating (coupled) mitochondrial respiration state.

Fig. 5

Hepatic β-hydroxybutyrate (HBA)-supported mitochondrial OXPHOS capacity is lower in individuals with steatosis. JO₂ values from the titration of HBA in the presence of saturating levels of ADP in (a) mouse and (b) human whole-liver, and (c) mouse and (d) human primary hepatocytes with non-linear regression fits based on the mean values according to the Michaelis Menten equation. JO₂ values from the titration of acetoacetate (ACA) in the presence of saturating levels of ADP and malate in (e) mouse and (f) human whole-liver, as well as (g) mouse and (h) human primary hepatocytes with non-linear regression fits based on the mean values according to the Michaelis Menten equation. Mitochondrial respiration (JO₂) in whole-liver from participants with and without hepatic steatosis (cut-off: histological hepatic fat content >5%), obtained with (i) HBA titrations (0, 1, 2.5, 5, 10, 15, 20 mM of HBA) in the presence of saturating levels of ADP and in the absence of malate and (j) the NADH (N) and succinate (S) combined (NS) pathway mitochondrial respiration protocol. (k) Relative contribution of the [HBA]_P respiratory state versus the maximal coupled N- and NS-linked mitochondrial respiration states. (l) Citrate Synthase (CS) activity and (m) protein expression of β-hydroxybutyrate dehydrogenase (BDH1). (n) Representative immunoblots of BDH1 and succinyl-CoA: 3-ketoacid (oxoacid) coenzyme A transferase (SCOT) protein expression from three replicates of each group are shown. Because SCOT is not expressed in liver tissue, a positive control was loaded on each gel to validate the suitability of the anti-SCOT antibody. a, c: n = 6. b, e, f, g: n = 7. d, h: n3. i, l, m: n = 5 versus 11, and j and k: n = 5 versus 10 for non-steatosis versus steatosis samples, respectively. Data are presented as mean ± SEM. a–h: Michaelis–Menten kinetic parameters (V_max and K_m values) were calculated based on the non-linear regression curve fitting and Michaelis Menten equation, and are expressed as JO₂ [pmol O₂ mg⁻¹ s⁻¹]. For statistical reasons (i.e., calculation of K_m values in GraphPad Prism), the maximal KB concentration shown in panels a to h is the one inducing maximal JO₂, despite titrations continuing up to 20–40 mM HBA and 15 mM ACA, depending on species, as reported in the Methods. i: Non-linear regression curves were fitted to the data points and differences between groups were determined using repeated-measurements ANOVA. j–k: repeated-measurements mixed-effects (REML) models with correction for multiple testing with the false discovery rate. l, m: Welch's t-test. D _E: electron transport chain capacity (noncoupled) mitochondrial respiration state; JO₂: oxygen flux; _L: leak mitochondrial respiration state; n/d: not determined; _P: phosphorylating (coupled) mitochondrial respiration state. Icons obtained from BioRender.com.

Fig. 6

Absolute and relative contribution of ketone body (KB)-driven mitochondrial ATP production in specific human and mouse organs in insulin resistant states. White arrows indicate direction of change in KB-linked mitochondrial oxidative phosphorylation (OXPHOS) capacity, as described in Fig. 2, Fig. 3, Fig. 4, Fig. 5. A downward arrow indicates a decrease. The solid and dotted lines in the rectangles in each quadrant represent the relative contribution of KB-driven OXPHOS capacity to maximal OXPHOS capacity, as determined in Fig. 2, Fig. 3, Fig. 4, Fig. 5 for control and diseased cohorts, respectively. Blue lines indicate human cohorts, grey lines indicate mouse cohorts. ACA: acetoacetate; AcCoA: acetyl coenzyme A; HBA: β-hydroxybutyrate; T2D: type 2 diabetes; TCA: tricarboxylic acid. Figure generated using BioRender.com.