High-Fat Diet Causes Mitochondrial Dysfunction as a Result of Impaired ADP Sensitivity

Although the etiology of type 2 diabetes is poorly defined, chronic consumption of a high-fat (HF) diet is a major contributor to whole-body glucose intolerance (1) and insulin resistance (2,3). The molecular explanation for these responses is not fully understood; however, mitochondrial dysfunction within skeletal muscle has received attention as a potential contributor because mitochondrial content is reduced in most reports of insulin-resistant/obese human skeletal muscle (4,5). Moreover, in various models, the induction of mitochondrial biogenesis protects against the development of insulin resistance (2,6), and genetic approaches that decrease mitochondrial content predispose animals to HF diet–induced insulin resistance (7). These data suggest changes in mitochondrial content are related to insulin sensitivity, but reductions in mitochondrial oxidative capacity are not uniformly reported in the literature with insulin resistance (8,9), and HF feeding has been shown to induce glucose intolerance and insulin resistance despite increasing mitochondrial content (10–12). Together, these data suggest reductions in mitochondrial respiratory capacity are not required for the development of insulin resistance.

Although short-term HF experiments challenge the notion that mitochondrial dysfunction contributes to insulin-resistance development, these data are difficult to rectify with molecular approaches that alter mitochondrial content and highlight a strong association with insulin sensitivity. An important consideration in this respect is that previous examinations of mitochondrial function have been performed in the presence of saturating ADP concentrations, which may not reflect the biological environment. An increase in mitochondrial content has been postulated to improve the sensitivity of oxidative metabolism to ADP (13,14), and therefore, ADP responsiveness may represent a key process in cellular homeostasis that is not recapitulated in experiments conducted in the presence or absence of saturating ADP. Importantly, the movement of ADP into the mitochondrial matrix and the subsequent binding of ADP to F₁F₀ ATP synthase decreases membrane potential and H₂O₂ production while simultaneously increasing substrate oxidation (15). In this manner, ADP sensitivity may represent a key biological process influenced by changing mitochondrial content because it can influence redox balance and reactive-lipid accumulation. Intriguingly, mitochondrial ADP sensitivity is externally regulated and inhibited by reactive-lipid accumulation (i.e., palmitoyl-CoA [P-CoA]); however, surprisingly, how mitochondrial ADP sensitivity is affected by HF consumption is not currently known. We hypothesized that mitochondrial ADP sensitivity would be impaired after an HF diet, independent of reductions in mitochondrial content. If accurate, this hypothesis would rectify a prominent discrepancy in the literature regarding the notion that mitochondrial dysfunction occurs in parallel with the development of insulin resistance.

All experimental procedures were approved by the University of Guelph Animal Care Committee and conformed to the Guide for the Care and Use of Laboratory Animals as indicated by the U.S. National Institutes of Health. Male mice (10–12 weeks of age) on a C57Bl6J background were fed a control (CON; 10% lard) or HF (60% lard) diet ad libitum for 4 weeks. All animals were anesthetized with an intraperitoneal injection of sodium pentobarbital (60 mg/kg; MTC Pharmaceuticals, Cambridge, Ontario, Canada) before red gastrocnemius muscle extraction for subsequent analyses.

Intraperitoneal glucose and insulin tolerance tests were performed on separate days as previously reported (16). On a separate day, animals were anesthetized, and the red gastrocnemius muscles were removed before or 15 min after an intravenous injection of 1 unit/kg body weight of insulin (NovoRapid). The area under the curve (AUC) calculations were adjusted to account for baseline blood glucose values.

Resting VO₂ and VCO₂ were monitored in metabolic cages (Columbus Instruments, Columbus, OH) and used to calculate total carbohydrate and fat oxidation and energy expenditure as previously reported (17).

Respiration in red gastrocnemius permeabilized muscle fibers was measured using high-resolution respirometry (Oxygraph-2k; Oroboros Instruments, Innsbruck, Austria) at 37°C as previously reported (13). Briefly, ADP (0–12,000 μmol/L) was titrated in the presence of pyruvate (10 mmol/L) and malate (5 mmol/L), followed by glutamate (10 mmol/L) and succinate (10 mmol/L), in the presence or absence of P-CoA (20 or 60 μmol/L). In separate experiments, P-CoA was titrated in the presence of l-carnitine (2 mmol/L), malate (5 mmol/L), and ADP (5 mmol/L) in the absence or presence of malonyl-CoA (M-CoA) (7 μmol/L). Carboxyatractyloside (0.2–1.6 μmol/L) was used to inhibit ADP-supported respiration to investigate changes in adenine nucleotide translocase (ANT) substrate sensitivity. Cytochrome c (10 μmol/L) did not stimulate respiration during experiments (data not shown). Succinate-supported (20 mmol/L) H₂O₂ emission was determined fluorometrically in the absence and presence of ADP (100 μmol/L) as previously reported (13).

Frozen red gastrocnemius (6–10 mg) was homogenized in Tris buffer (100 mmol/L, pH 8.3) and freeze-thawed to lyse mitochondria. Citrate synthase activity was measured spectrophotometrically as previously reported (18).

The red gastrocnemius was homogenized, and Western blotting was performed as previously reported (17). See Supplementary Fig. 1 for a complete list of antibodies.

Michaelis-Menten kinetics were determined by plotting data points in GraphPad Prism 5 software to estimate the apparent K_m as previously described (19). Unpaired two-tailed Student t tests were used to analyze data between CON- and HF-fed mice, and one-way ANOVAs were used for P-CoA–ADP sensitivity comparisons, followed by Student-Newman-Keuls post hoc analyses where appropriate. Statistical significance was determined at P < 0.05. Data are expressed as mean ± SEM.

We first verified the expected HF diet responses on body mass and insulin-resistance. HF consumption resulted in greater body weight (CON: 31 ± 1.78 g, HF: 38.6 ± 0.95 g; P < 0.05), greater epididymal fat mass (CON: 1.2 ± 0.24 g, HF: 2.9 ± 0.15 g; P < 0.05), greater AUC in response to a glucose tolerance test and insulin tolerance test (Fig. 1A), and lower skeletal muscle insulin–stimulated AKT phosphorylation (Fig. 1B). Moreover, although mice exhibited the expected diurnal changes in fuel metabolism (Fig. 1C–F), consumption of the HF diet resulted in lower carbohydrate oxidation (Fig. 1C), higher fat oxidation (Fig. 1D), lower respiratory exchange ratio (Fig. 1E), and greater energy expenditure (Fig. 1F), further confirming the HF diet model.

With regard to potential mitochondrial changes influenced by HF consumption, we first examined markers of mitochondrial content. HF consumption increased peroxisome proliferator–activated receptor γ coactivator (PGC-1α) and several markers of the electron transport chain, total electron transport chain abundance, and pyruvate dehydrogenase E1α (PDHE1α), without altering carnitine palmitoyltransferase I (CPT-I), ATP synthase, and a subunit of cytochrome c oxidase (Fig. 2A). Moreover, HF consumption increased citrate synthase activity in support of a greater oxidative capacity (Fig. 2B). Despite the apparent HF diet–induced mitochondrial biogenesis, HF consumption did not alter state 2 respiration, maximal ADP-stimulated respiration, or respiratory control ratio values (Fig. 2C). Similarly, HF consumption did not affect maximal CPT-I–dependent P-CoA–supported respiration (Fig. 2D) or P-CoA sensitivity (Fig. 2E). Moreover, the CPT-I inhibitor M-CoA attenuated P-CoA sensitivity, but this was not affected by HF consumption (Fig. 2F). Although these data strongly suggest the absence of mitochondrial respiratory dysfunction after HF consumption, given that skeletal muscle contains submaximal ADP concentrations, we next examined mitochondrial respiration at physiological concentrations and ADP sensitivity.

Given the absence of changes in maximal ADP-stimulated respiration, we next examined mitochondrial respiration across a range of ADP concentrations. The HF diet did not impair respiration in the presence of >2,000 μmol/L ADP, but strikingly, respiration was attenuated ∼30% at all biologically relevant ADP concentrations (i.e., 100–1,000 μmol/L ADP) (Fig. 3A). In addition, mitochondrial ADP sensitivity was decreased ∼25% after HF consumption (i.e., greater apparent K_m value) (Fig. 3B). Combined, despite the induction of mitochondrial biogenesis, these data strongly suggest the development of mitochondrial dysfunction with an HF diet secondary to impaired ADP sensitivity.

Given that ADP binding to F₁F₀ ATP synthase is known to suppress H₂O₂ emission rates (15), we also examined mitochondrial H₂O₂ emissions in the absence and presence of ADP to further solidify this relationship. The maximal capacity for H₂O₂ emission did not change after the HF diet (Fig. 3C), but mice fed the HF diet had an impaired ability to suppress H₂O₂ emission rates via ADP transport into the mitochondria, as indicated by H₂O₂ production in the presence of ADP that was approaching significance (Fig. 3D) and a lower percentage suppression of H₂O₂ emission rates by ADP (Fig. 3E). Further, mice fed the HF diet had greater antioxidant expression (catalase and superoxide dismutase 2) and 4-hydroxynonenal adducts than mice fed the CON diet (Fig. 3F), supporting an increase in oxidative stress/damage and stimulation of antioxidant adaptation during HF consumption. The impairment in ADP sensitivity occurred independent of changes in protein expression of ADP transporters (i.e., voltage-dependent anion channel [VDAC] and ANT), mitochondrial creatine kinase, cytosolic creatine kinase, and uncoupling protein 3 (Fig. 3F). We therefore examined the ability of carboxyatractyloside to inhibit ADP-stimulated respiration to gain insight into the possibility that posttranslational modifications on ANT contribute to the attenuated ADP responsiveness after the HF diet. Although 1.6 μmol/L carboxyatractyloside inhibited respiration ∼90% regardless of diet (data not shown), respiration was higher after the HF diet in the presence of 0.2 μmol/L carboxyatractyloside (Fig. 3G), suggesting a decreased ability for carboxyatractyloside to interact with the ADP-binding motif on ANT. Altogether, these data suggest an HF diet attenuates ANT sensitivity to ADP, likely contributing to the apparent induction in mitochondrial dysfunction.

Although these data strongly suggest impaired ADP responsiveness after the HF diet, these experiments were conducted in optimal conditions. However, long-chain fatty acid CoAs (e.g., P-CoA) are known to increase after HF consumption (20), and this lipid derivative has been shown to inhibit ANT and reduce ADP sensitivity in permeabilized muscle fibers (13,17). Therefore, given the insensitivity to carboxyatractyloside, we next examined P-CoA–mediated inhibition of ADP sensitivity in CON- and HF-fed mice. To do this, we repeated experiments on mitochondrial ADP sensitivity in the absence and presence of P-CoA concentrations that better reflect CON (20 µmol/L) (21) and HF diet intramuscular situations (60 µmol/L) (20). P-CoA concentrations that reflected CON conditions did not alter respiration rates at any ADP concentration (data not shown), but the presence of 60 µmol/L P-CoA attenuated respiration at most ADP concentrations (Fig. 4A) and further attenuated ADP sensitivity after the HF diet (i.e., increased the apparent ADP K_m) (Fig. 4B). Combined, these data indicate the presence of mitochondrial respiratory dysfunction after the HF diet is exacerbated in the presence of increased intramuscular P-CoA concentrations.

We provide compelling evidence that consumption of an HF diet induced mitochondrial dysfunction as a result of impaired ADP sensitivity. Specifically, we show that despite the induction of mitochondrial biogenesis and unaltered maximal ADP-stimulated respiration, respiration at physiological ADP concentrations was impaired after the HF diet. In addition, the ability of ADP to stimulate respiration (apparent K_m) and attenuate mitochondrial H₂O₂ emission were decreased after the HF diet, whereas modeling reactive lipid levels that occur with HF consumption exacerbated the apparent mitochondrial dysfunction. Combined, these data strongly indicate impaired mitochondrial bioenergetics after an HF diet due to impaired ADP sensitivity.

Whereas genetic models and interventions that increase mitochondrial content typically have improved insulin sensitivity and glucose tolerance (1,2,6), consumption of the HF diet in the current study, and in others (10,11), paradoxically resulted in increased mitochondrial content despite the induction of glucose intolerance and insulin resistance. Therefore, the notion that mitochondrial dysfunction contributes to the development of insulin resistance has waned in recent years. Our data, however, implicate HF consumption in impairing respiration and increasing mitochondrial H₂O₂ emission as a result of impaired ADP sensitivity. These derangements appear to be amplified in the presence of higher P-CoA concentrations indicative of insulin-resistant muscle, suggesting the presence of mitochondrial ADP insensitivity in skeletal muscle after HF consumption and insulin resistance. The constant VDAC/ANT and ATP synthase protein abundance suggest altered regulation of these proteins may lead to mitochondrial dysfunction. Whereas an impairment of ATP synthase may also exist after an HF diet, the insensitivity to carboxyatractyloside suggests that ANT is less sensitive to ADP binding, a response likely amplified by increases in intramuscular P-CoA concentrations (22,23).

Our data show that impaired mitochondrial ADP sensitivity plays an important role in HF diet–induced mitochondrial dysfunction. Although speculative, the induction of mitochondrial biogenesis likely represents a compensatory mechanism to improve ADP sensitivity and potentially mitigate HF diet–induced redox stress (24); however, this appears inadequate to maintain cellular homeostasis. Whether the reduction in ADP sensitivity directly contributes to the metabolic inflexibility observed in HF mice is unknown. However, it is also possible that this response, similar to the induction of mitochondrial biogenesis, represents a compensatory adaptation to preserve some carbohydrate utilization in the presence of increased intramuscular lipid availability, because a rise in cytosolic ADP could in theory promote carbohydrate metabolism. In addition to the potential influence on fuel selection, because the movement of ADP into the mitochondrial matrix can bind to ATP synthase to stimulate respiration and attenuate mitochondrial-derived H₂O₂ emission (15), an impairment in ADP transport may represent a nexus point influencing several models implicated in the development of insulin resistance.

Funding. P.M.M. is supported by a Natural Sciences and Engineering Research Council of Canada graduate scholarship. This work was supported by the Natural Sciences and Engineering Research Council of Canada (03656 to G.P.H.). Infrastructure was purchased with assistance from the Canadian Foundation for Innovation/Ontario Research Fund.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. P.M.M. conducted experiments, analyzed data, and drafted the manuscript. P.M.M., P.J.L., and G.P.H. interpreted results and edited the manuscript. P.M.M. and G.P.H. designed the study. G.P.H. is guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.