Author + information
- Received May 8, 2009
- Revision received June 24, 2009
- Accepted July 6, 2009
- Published online November 10, 2009.
- Ethan J. Anderson, PhD⁎,†,‡,⁎ (, )
- Alan P. Kypson, MD⁎,
- Evelio Rodriguez, MD⁎,
- Curtis A. Anderson, MD⁎,
- Eric J. Lehr, MD, PhD⁎ and
- P. Darrell Neufer, PhD†,‡
- ↵⁎Reprint requests and correspondence:
Dr. Ethan J. Anderson, Department of Cardiovascular Sciences, East Carolina Heart Institute, East Carolina University, Heart Drive, Greenville, North Carolina 27835
Objectives The aim of this study was to determine the impact of diabetes on oxidant balance and mitochondrial metabolism of carbohydrate- and lipid-based substrates in myocardium of type 2 diabetic patients.
Background Heart failure represents a major cause of death among diabetic patients. It has been proposed that derangements in cardiac metabolism and oxidative stress may underlie the progression of this comorbidity, but scarce evidence exists in support of this mechanism in humans.
Methods Mitochondrial oxygen (O2) consumption and hydrogen peroxide (H2O2) emission were measured in permeabilized myofibers prepared from samples of the right atrial appendage obtained from nondiabetic (n = 13) and diabetic (n = 11) patients undergoing nonemergent coronary artery bypass graft surgery.
Results Mitochondria in atrial tissue of type 2 diabetic individuals show a sharply decreased capacity for glutamate and fatty acid-supported respiration, in addition to an increased content of myocardial triglycerides, as compared to nondiabetic patients. Furthermore, diabetic patients show an increased mitochondrial H2O2emission during oxidation of carbohydrate- and lipid-based substrates, depleted glutathione, and evidence of persistent oxidative stress in their atrial tissue.
Conclusions These findings are the first to directly investigate the effects of type 2 diabetes on a panoply of mitochondrial functions in the human myocardium using cellular and molecular approaches, and they show that mitochondria in diabetic human hearts have specific impairments in maximal capacity to oxidize fatty acids and glutamate, yet increased mitochondrial H2O2emission, providing insight into the role of mitochondrial dysfunction and oxidative stress in the pathogenesis of heart failure in diabetic patients.
The cascade of metabolic events that progressively leads to type 2 diabetes consists of early hyperlipidemia and hyperinsulinemia, followed eventually by beta-cell demise and hyperglycemia, the latter defining the disease. Each of these metabolic perturbations is thought to contribute individually, but also collectively, to altered cellular structure and electromechanical function in the diabetic myocardium, a condition known clinically as diabetic cardiomyopathy (1). In Western societies, the rapidly increasing number of type 2 diabetic patients, coupled with the obesity epidemic, illustrates the need for studies that specifically focus on addressing the cellular and molecular mechanisms driving the pathology of this comorbidity.
In diabetes, the elevated levels of serum triglycerides and free fatty acids result in pronounced accumulation of myocardial triglycerides, a phenomenon that has been well established in experimental models (2,3) and humans (4,5). This unbalanced lipid metabolism leads to cardiac steatosis, a condition proposed to play a causative role in the development of contractile dysfunction in the diabetic human myocardium. A decreased ratio of adenine triphosphate produced per oxygen (O2) consumed (measure of mitochondrial efficiency) is also evident in the diabetic myocardium, raising the possibility that mitochondrial dysfunction may be an underlying cause of the cardiomyopathy (6). Furthermore, increased mitochondrial reactive oxygen species (ROS) production has been shown to accompany this mitochondrial dysfunction (7).
To date, the lack of approaches to investigate metabolism of carbohydrate- and lipid-based substrates at the subcellular level (e.g., mitochondria) in human myocardium represents a significant obstacle to identifying the mechanisms responsible for myopathy in the diabetic heart. Recently, the use of permeabilized myofibers as an in vitro model of mitochondrial function has provided a number of mechanistic insights concerning the role of mitochondria in diseases affecting human skeletal (8,9) and cardiac (10) muscle. In this study we have used this system in a biochemical approach to investigate the effects of type 2 diabetes on mitochondrial respiration and oxidant emission under a diverse range of substrate conditions, as well as the global redox environment in atrial appendage tissue from patients undergoing nonemergent coronary artery bypass graft (CABG) surgery.
Patient demographics and clinical characteristics
Approval for this study was granted by the institutional review board of East Carolina University. Informed consent was obtained from patients at Pitt County Memorial Hospital undergoing CABG using cardiopulmonary bypass and hypothermic cardioplegic arrest. All demographic and clinical data pertaining to the patients who participated in this study are shown in Table 1.The patients were grouped as either nondiabetic or diabetic according to 2 major variables: 1) clinical diagnosis of diabetes; and 2) glycosylated hemoglobin (HbA1c) values of ≥6.1 extending up to approximately 1 year before surgery. The vast majority of diabetic patients was given intravenous insulin for ≥48 h before the procedure (standard of care). Patients with enlarged atria, a history of arrhythmia, or left ventricular ejection fractions ≤30% were excluded from this study.
Human atrial appendage biopsy and tissue processing
After median sternotomy, and before institution of cardiopulmonary bypass, a purse-string suture was placed in the right atrial appendage to allow for placement of the venous cannula. A sample of the appendage directly superior to the purse-string suture was dissected and immediately rinsed in ice-cold Buffer X (11). This process results in a pristine human myocardial specimen that has not been subjected to potentially confounding variables such as cardioplegic arrest, mechanical handling, or contact with tubing. The sample was then blotted on gauze to remove excess buffer and was trimmed of the epicardial layer and pericardial fat, and a small portion was immediately frozen in liquid N2 for protein and mRNA analysis.
Permeabilized muscle fiber preparation; mitochondrial O2and hydrogen peroxide (H2O2) measurements
The technique of permeabilized fiber bundle preparation is adapted from previous methods (11,12). Mitochondrial O2consumption measurements were made using the Oroboros O2K Oxygraph (Innsbruck, Austria), and mitochondrial H2O2emission was measured using a Horiba Jobin Yvon spectrofluorometer (Edison, New Jersey) according to methods described previously (9).
Atrial tissue glutathione (reduced glutathione [GSH] and oxidized glutathione [GSSG]) and triglyceride measurements
Atrial muscle samples frozen in liquid N2 were pulverized and homogenized, and protein samples were prepared for glutathione measurements as described previously (9). Total GSH and GSSG were measured using the reagents and calibration set provided by the GSH/GSSG assay (Oxis Research, Beverly Hills, California) according to the manufacturer's instructions, with some small modifications. Triglycerides were measured in atrial muscle homogenate using a colorimetric assay kit provided by BioVision (Mountain View, California), according to the manufacturer's instructions.
Immunoblotting and protein quantification
Samples of atrial muscle protein homogenate were separated using sodium dodecyl sulfate polyacrylamide gel electrophoresis, transferred to polyvinylidene fluoride membranes, and subjected to immunoblot using antibodies for peroxisome proliferator-activated receptor (PPAR)-alpha (Abcam, Cambridge, Massachusetts), peroxisome proliferator-activated receptor gamma coactivator-1 (PGC1)-alpha (Cell Signaling Technology, Danvers, Massachusetts), 3-nitrotyrosine (Abcam), and hydroxynonenal (HNE) adduct (Oxis Research). Densitometric analysis was performed using Image J software (13).
All statistical and graphical analysis was performed using GraphPad Prism version 5.0 (GraphPad software, San Diego, California). Categorical variables were compared using the Fisher exact test, and all interval variables were compared using the Student unpaired ttest. Data shown in Table 1are expressed as mean ± SD. All experimental data are expressed as mean ± SEM. Differences between nondiabetic patients and diabetic patients were considered statistically significant for p < 0.05. Regression analysis was used to examine the correlation between glycosylated hemoglobin (HbA1c) and myocardial triglycerides and maximal rate of fatty acid oxidation in both groups. For measurements of kinetics of mitochondrial O2consumption and H2O2emission in each group, best-fit curves were obtained using nonlinear regression analysis, and statistically significant differences between groups were confirmed by comparison of the R2values.
Triglyceride levels and mitochondrial fatty acid–supported respiration
Atrial tissue obtained from diabetic patients undergoing CABG contained an approximately 2-fold greater level of intramyocellular lipid (IMCL) triglyceride content than that from nondiabetic patients (Fig. 1A),which positively correlated with HbA1cin these patients (Fig. 1B). To determine whether this increased IMCL triglyceride might be linked to reduced mitochondrial oxidation of fatty acids in the atrium of diabetic patients, we investigated the kinetics of O2consumption supported by palmitoyl-L-carnitine (an activated fatty acid) in permeabilized atrial myofibers prepared from diabetic and nondiabetic patients. A glucose/hexokinase adenosine diphosphate (ADP)–regenerative system was used to maintain a continuous ADP-stimulated respiratory state to ensure that substrate oxidation was not impeded by thermodynamic constraints. By titrating palmitoyl-L-carnitine during continuous maximal state-3 respiration, we were able to measure both the sensitivity of mitochondrial transporters and beta-oxidation enzymes for palmitoyl-L-carnitine (K0.5) as well as its maximal oxidation rate (Vmax). Although the K0.5for palmitoyl-L-carnitine was unchanged, Vmaxsupported by palmitoyl-L-carnitine was significantly reduced in permeabilized atrial fibers from diabetic compared with nondiabetic patients (Fig. 1D). This reduced maximal fatty acid oxidation capacity in atrial fibers was negatively correlated with levels of blood HbA1c(Fig. 1E). Citrate synthase activity was similar in atrial tissue homogenates between groups (30.8 ± 3.3 μmol·min−1·mg−1vs. 29.8 ± 2.4 μmol·min−1·mg−1protein in nondiabetic vs. diabetic subjects), providing evidence that the reduced fatty acid oxidative capacity in the atrial tissue of diabetic patients is not attributable to an overall reduction in mitochondrial content.
Expression of PPAR-α and PGC1α
As a nuclear receptor responsible for determining substrate preference in the heart, PPARα coordinates expression of most key regulators of fatty acid metabolism (14). Because palmitoyl-L-carnitine supported respiration was so strongly diminished in diabetic atrial tissue, we postulated that this might be caused by altered expression of PPAR-α or its transcriptional coactivator PGC1α, a crucial transcriptional co-activator of numerous mitochondrial genes. Although the mean PPAR-α protein levels were slightly lower, and PGC1α was slightly higher in diabetic atrial tissue, the difference did not reach statistical significance in this cohort of patients (data not shown), providing evidence that the reduced palmitoyl-carnitine respiration observed in the diabetic atrium is not caused by reduced expression of these key regulatory proteins.
Respiration supported by tricarboxylic acid cycle substrates and kinetics of ADP-stimulated respiration
Next, we examined whether respiratory control and/or capacity during respiration supported by carbohydrate-based substrates was altered in atrial tissue prepared from diabetic versus nondiabetic patients. Titration of ADP during respiration supported by maximal pyruvate (Fig. 2A)or succinate (Fig. 2B) showed similar submaximal and maximal respiratory kinetics, indicating that activities of the pyruvate dehydrogenase, succinate dehydrogenase (complex II of the respiratory system), and adenine nucleotide translocase (a key regulator of energy transfer in cardiomyocytes [15,16]) are not different in atrial muscle from diabetic patients as compared with that from nondiabetic patients.
We then asked whether maximal respiration supported by glutamate was altered in the atrial tissue of diabetic patients, because glutamate oxidation provides a significant source of fuel to the mammalian heart (17,18). Maximal glutamate oxidation was reduced in permeabilized atrial myofibers prepared from diabetic patients, as assessed by respiration in the presence of glutamate and malate. This was not caused by disruption of the outer mitochondrial membrane and/or cytochrome C deficiency, because the addition of exogenous cytochrome C failed to restore respiration in diabetic atrial myofibers (Figs. 2C and 2D).
Mitochondrial H2O2emission and oxidative stress
Because a previous report (19) has shown oxidative damage to mitochondria in diabetic mononuclear cells, we sought to determine whether the propensity for mitochondria to emit oxidants (i.e., mitochondrial oxidant-emitting potential ) is elevated in atrial tissue from patients with type 2 diabetes. First, we measured H2O2emission in the presence of incrementally increasing concentrations of succinate (with oligomycin present to inhibit mitochondrial adenosine triphosphatase) in permeabilized atrial myofibers prepared from diabetic and nondiabetic patients. Mitochondria in diabetic atrial tissue showed both a greater maximal rate of H2O2emission supported by succinate as well as a greater rate of H2O2emission at low succinate concentrations (Fig. 3A),indicating a greater mitochondrial oxidant-emitting potential in the myocardium of these patients. This increased H2O2emission with succinate was not caused by increased flux through complex II (i.e., increased succinate dehydrogenase activity) because parallel experiments showed similar rates of O2consumption (Fig. 3B).
To determine the mitochondrial oxidant-emitting potential of atrial fibers under conditions more representative of the physiological state, we measured H2O2emission during continuous submaximal state 3 respiration (i.e., using a glucose/hexokinase ADP regenerative system) supported by palmitoyl-L-carnitine alone and after addition of glutamate, malate, and succinate to simulate oxidation of substrates from multiple sources. Mitochondria in diabetic atrial tissue showed greater rates of H2O2emission than in nondiabetic patients under all substrate conditions examined (Fig. 3C). Surprisingly, a greater rate of mitochondrial H2O2emission in diabetic atrial tissue with palmitoyl-L-carnitine was evident despite the decreased capacity for palmitoyl-L-carnitine supported respiration, which translates to a much greater ratio of moles of H2O2emitted per mole of O2consumed with this substrate (Fig. 3D).
As a result of the presence of increased mitochondrial H2O2emission in diabetic atrial tissue, we speculated that markers of oxidative stress would be present in this tissue. Atrial tissue from diabetic patients showed both a greater concentration of GSSG and a reduced concentration of total glutathione than that found in nondiabetic patients (Fig. 4A),which corresponded to a decreased GSH/GSSG ratio in the diabetic patients (Fig. 4B). Higher steady-state levels of lipid peroxidation and nitrosative stress were also detected in diabetic atrial tissue, as assessed by immunoblot analysis of proteins from atrial homogenate using antibodies that react with HNE- and 3-nitrotyrosine–modified proteins, respectively (Figs. 4C to 4F).
By using a permeabilized fiber approach on atrial tissue obtained during coronary bypass surgery, the present study is the first to comprehensively address the effect of type 2 diabetes on mitochondrial metabolism of lipid- and carbohydrate-based substrates and ROS emission in human myocardium, providing evidence that: 1) maximal capacity for mitochondrial oxidation of palmitoyl-carnitine is decreased in the atrium of type 2 diabetic human myocardium; 2) mitochondrial content and expression of PPAR-α and PGC1α are unchanged in the atrium of a diabetic patient; 3) glutamate but not pyruvate or succinate oxidation is decreased in the atrium of a diabetic patient; and 4) mitochondrial H2O2emission supported by both carbohydrate- and lipid-based substrates is greater in the atrium of a diabetic patient, corresponding to increased oxidative stress in this tissue. These findings show that mitochondria in the diabetic human heart have specific impairments that limit the maximal capacity to oxidize palmitoyl-carnitine and glutamate. Furthermore, they suggest that despite these impairments, the propensity for the electron transport system to emit oxidants is elevated in the diabetic mitochondria, likely contributing to increased oxidative stress and the marked decline in cardiac electromechanical function that is known to occur in diabetic myocardium over time, ultimately leading to heart failure.
Cardiac steatosis in diabetes
Hyperlipidemia, a prominent characteristic of type 2 diabetes mellitus, has been shown to cause a decrease in glucose and lactate use and an increase in fatty acid uptake and oxidation by the myocardium (1,20,21). However, in diabetes, the increased levels of serum free fatty acids and triglycerides often exceed their demand in the working heart, leading to a buildup of IMCL triglycerides (i.e., steatosis) in the tissue (2,3). In the present study, the IMCL triglyceride content was significantly higher in diabetic human atrial tissue, likely a consequence of the elevated serum triglyceride levels, supporting the concept of a chronic mismatch between substrate supply and metabolic demand. In addition, the maximal capacity for mitochondrial fatty acid oxidation was markedly lower in diabetic atrial fibers, implying a maladaptive response or impairment at the level of beta-oxidation and/or the respiratory system. Whether a reduced capacity to oxidize fatty acids is present throughout the heart, particularly in the left ventricles (LV) of these patients, remains to be determined, although an exquisite series of recent studies using noninvasive spectroscopic techniques has shown the presence of elevated IMCL triglyceride in the LV of obese and type 2 diabetic patients (4,5). Importantly, the increased IMCL triglyceride content was found to strongly correlate with diastolic dysfunction in type 2 diabetic patients, independent of BMI, age, and hypertension (5). These investigators hypothesized that as a result of LV steatosis, diastolic dysfunction manifests early on in the disease, progressively leading to global contractile dysfunction (i.e., systolic dysfunction) in the LV with time. The role of steatosis in contractile dysfunction is controversial, however. Recently, intriguing studies have suggested that accumulation of triglycerides in both skeletal (22) and cardiac (23) muscle as a result of diet-induced obesity is not at all pathogenic, but may even be protective against obesity-associated maladies such as insulin resistance, implying that alternative explanations for the association between cardiac steatosis and diabetic cardiomyopathy should be explored. We and others (11) have shown that activated fatty acids generate a substantial amount of mitochondrial ROS, representing another possible route by which adverse effects are generated in the diabetic myocardium.
Mitochondrial fatty acid oxidation in diabetic myocardium
In the present study, the finding that maximal fatty acid–supported respiration is decreased in atrial tissue from diabetic patients is somewhat surprising and unexpected in light of a recent study by Herrero et al. (24), which found that myocardial fatty acid oxidation was elevated in diabetic patients in vivo. However, that study was conducted on young (25 to 35 years old) patients in a precisely controlled post-prandial state, whereas in the present study, all patients were much older (45 to 65 years of age) and fasted for a prolonged period of time (approximately 12 to 18 h) before surgery (i.e., when the biopsy was taken). It is conceivable that diabetic myocardium preferentially oxidizes lipids to a greater extent than that of nondiabetic patients in the post-prandial state, but then is unable to appropriately increase capacity for lipid oxidation during the transition to the fasted state. This is similar to the observation by Young et al. (25), which showed that the myocardium in the obese, insulin-resistant Zucker rat was unable to appropriately up-regulate fatty acid oxidation to levels comparable to those of lean rats in response to fasting. In addition, myocardial insulin resistance is a plausible explanation for the impaired palmitoyl-carnitine and glutamate oxidation evident in the diabetic atrial tissue, as a recent study by Boudina et al. (26) showed that cardiac mitochondria in mice with a cardiac-specific deletion of the insulin receptor showed a strikingly similar phenotype.
Mitochondrial ROS and oxidative stress in diabetic myocardium
In the present study, mitochondria in diabetic human atrial tissue showed much greater rates of H2O2emission than those in tissue of nondiabetic patients, regardless of the substrate used or respiratory state. These findings suggest that alterations in the electron transport system and/or antioxidant capacity are present in the diabetic human myocardium and provide evidence that mitochondria may be a significant source of the ROS in this tissue.
Moreover, persistent oxidative stress is evident in diabetic human atrial tissue, as shown by the increased levels of HNE- and 3-nitrotyrosine–modified proteins, which can only occur as a result of constitutively increased ROS production (or decreased removal) in the tissue. Interestingly, both HNE and tyrosine nitrosylation have been shown to alter specific proteins involved in metabolism in the mitochondrial inner membrane and matrix in a manner that alters their activity (27).
Clinical relevance of the present study
Collectively, the findings of the present study show that mitochondria in the diabetic human heart have specific impairments in maximal capacity to oxidize fatty acids and glutamate, in addition to an increased propensity for mitochondrial H2O2emission and evidence of persistent oxidative stress, thus providing important insight on the effects of type 2 diabetes in the human myocardium. Because this study was performed on atrial tissue, an important follow-up study would be to examine whether the same events occur in ventricular tissue, given the well-established anatomic and physiological differences between these 2 myocardial compartments. Because ventricular tissue is extremely difficult to obtain from a viable human heart, studies in large-animal models of diabetes that more closely mimic the human condition (e.g., use of drugs for glycemic and lipidemic control, aging models) should be considered to investigate more precisely the effects of diabetes on myocardial metabolism in a translational manner. Ultimately, this could lead to better therapeutic approaches and proper evaluation of diabetic patients in the future.
The authors thank Robert Lust for helpful discussions and T. Bruce Ferguson and W. Randolph Chitwood for kind support on this project.
This work was supported by grants DK073488 and DK074825 from the National Institutes of Health.
- Abbreviations and Acronyms
- adenine diphosphate
- coronary artery bypass graft
- reduced glutathione
- oxidized glutathione
- glycosylated hemoglobin
- intramyocellular lipid
- left ventricle/ventricular
- peroxisome proliferator-activated receptor gamma coactivator-1
- peroxisome proliferator-activated receptor
- reactive oxygen species
- Received May 8, 2009.
- Revision received June 24, 2009.
- Accepted July 6, 2009.
- American College of Cardiology Foundation
- Boudina S.,
- Abel E.D.
- Zhou Y.T.,
- Grayburn P.,
- Karim A.,
- et al.
- McGavock J.M.,
- Lingvay I.,
- Zib I.,
- et al.
- Rijzewijk L.J.,
- van der Meer R.W.,
- Smit J.W.,
- et al.
- How O.J.,
- Aasum E.,
- Severson D.L.,
- Chan W.Y.,
- Essop M.F.,
- Larsen T.S.
- Boudina S.,
- Sena S.,
- Theobald H.,
- et al.
- Garnier A.,
- Fortin D.,
- Zoll J.,
- et al.
- Anderson E.J.,
- Yamazaki H.,
- Neufer P.D.
- Anderson E.J.,
- Neufer P.D.
- Rasband W.S.
- Ussher J.R.,
- Koves T.R.,
- Jaswal J.S.,
- et al.
- Herrero P.,
- Peterson L.R.,
- McGill J.B.,
- et al.
- Young M.E.,
- Guthrie P.H.,
- Razeghi P.,
- et al.
- Boudina S.,
- Bugger H.,
- Sena S.,
- et al.