Bioenergetic functions in subpopulations of heart mitochondria are preserved in a non-obese type 2 diabetes rat model (Goto-Kakizaki)

A distinct bioenergetic impairment of heart mitochondrial subpopulations in diabetic cardiomyopathy is associated with obesity; however, many type 2 diabetic (T2DM) patients with high-risk for cardiovascular disease are not obese. In the absence of obesity, it is unclear whether bioenergetic function in the subpopulations of mitochondria is affected in heart with T2DM. To address this issue, a rat model of non-obese T2DM was used to study heart mitochondrial energy metabolism, measuring bioenergetics and enzyme activities of the electron transport chain (ETC). Oxidative phosphorylation in the presence of substrates for ETC and ETC activities in both populations of heart mitochondria in T2DM rats were unchanged. Despite the preservation of mitochondrial function, aconitase activity in T2DM heart was reduced, suggesting oxidative stress in mitochondria. Our study indicate that metabolic function of heart mitochondria is unchanged in the face of oxidative stress and point to a critical role of obesity in T2DM cardiomyopathy.


Results
Animal model. The animal model characteristics are reported in Table 1. The body and heart weight of diabetic (GK) rats is significantly reduced in comparison to the control (W) rats at both 18 and 28 weeks. The body and heart weight of GK rats does not change from 18 to 28 weeks, while those of W rats significantly increases by 24% 23 and 28%, respectively. The GK rats are hyper-insulinemic and hyperglycemic at 18 and 28 weeks.
For both group of rats, the insulin tolerance test with the time profile of blood glucose concentrations is reported in Fig. 1. Blood glucose concentration is normalized to the basal blood glucose concentration measured before insulin injection. In the first 20 minutes of the test, the normalized blood glucose content in GK was significantly higher than that observed for W. The basal glucose concentration of W and GK was 7.5 ± 0.6 and 12.6 ± 0.4 mM.  23 and heart weight, insulin and glucose concentrations 23 in blood. Data are mean ± SD (n = 6). Influence of insulin resistance within same age (p < 10 −3 ): (*) Statistically different from control. Influence of age within group (p < 10 −5 ): ( # ) Statistically different from 18 weeks. Mitochondrial yield and enzymes. The yield of SSM and IFM did not differ in GK and W rats or with age (Table 2). CS and SDH activities were similar in GK and W rats at 18 and 28 wk. At 18 wk, the activity of aconitase was significantly reduced in both subpopulations of mitochondria of GK rats in comparison to the control group (Table 2), whereas at 28 wk in IFM, it was significantly lower in GK than W; there is a trend toward significance between SSM GK and W rats.
immunoblotting. To confirm that the expression of mitochondrial aconitase was unaltered in our T2DM lean diabetic model, we determined ACO-2 (aconitase) content by immunoblotting in subpopulation of mitochondria isolated from rat hearts at 28w (Fig. 2). In both SSM and IFM the ACO-2 expression was not different in GK rats as compared to W rats, despite a tendency to be higher in GK SSM, albeit not statistically significant.
oxidative phosphorylation. Oxidative phosphorylation for heart muscle SSM and IFM was measured using substrates for complex I (glutamate Table 3), I, II, III, and IV in both GK and W rats (Figs. 3 and 4) at 18 and 28 weeks. The state 3 and 4 respiratory rates in SSM were similar in the GK rats compared to those in the W (Table 3). Also, in GK and control rats at both ages, SSM respiration rates obtained with a saturated concentration of ADP increased ~20% and were comparable (Fig. 4) Table 2. Yields and enzyme activities of subsarcolemmal (SSM) and interfibrillar (IFM) mitochondria. a The mitochondrial yields and enzymatic activity of the subpopulations of heart mitochondria is normalized to gram of wet weight of heart muscle (gww) and mitochondrial protein (mg), respectively. Data are mean ± SD (n = 6). Influence of insulin resistance within same age (p < 5 10  saturating. With the addition of an uncoupler, the respiratory rates did not increase, indicating that oxidative phosphorylation is limited by oxidation in both groups of rats. In the IFM of both groups of rats, the results were similar to those obtained for SSM. Thus, in this model of T2DM oxidative phosphorylation rates were unaffected by insulin resistance (Figs. 3 and 4). The state 3 respiration rate obtained with saturated concentration of ADP (i.e. High ADP, Table 3), was significantly correlated to the activity of CS (r 2 = 0.63, p < 10 −5 ) or SDH (r 2 = 0.65, p < 10 −5 ) with a slope of the linear relationship significantly different from zero in both cases. The RCR of both groups was higher than 10, indicating that both populations of mitochondria are highly coupled. The ADP/O ratio is the same for both populations of   www.nature.com/scientificreports www.nature.com/scientificreports/ mitochondria ( Table 3). The ADP/O ratios determined for complexes I, II, and III substrates were similar in GK and W rats for both subpopulations of mitochondria ( electron transport chain. The activity of the ETC complexes was evaluated with specific spectrophotometric assays. In SSM and IFM of GK rats, the activity of the ETC components of a) complex I, II, III, and IV; b) linked complex I and III (NCR), c) flavin protein domain of complex I (NFR); d) linked complex II and III (SCR); e) complex II (SDH) were similar to those of the W (Fig. 6) at 18 and 28 wk.

Discussion
This study focused on subpopulations of heart mitochondria of GK rats to compare their bioenergetics with that of control rats at 18 and 28 weeks. Heart mitochondria were obtained from the same rats investigated in a previous study on skeletal muscle mitochondria 23 .
In SSM and IFM heart mitochondria of non-obese diabetic rats (GK) bioenergetic function was similar to the control group both at the age of 18 and 28 weeks. This was true with substrates for complex I, II, III, and IV, as well as of fatty acid oxidation. Electron transport chain activities were unchanged in GK confirming that mitochondrial bioenergetic function was preserved. The reduced aconitase activity in GK heart mitochondrial subpopulations indicates that despite the presence of oxidative stress, the bioenergetic function was preserved.
Animal model. In our work, GK rats were resistant to a decrease in blood glucose level after a bolus injection of insulin (Fig. 1). The delayed effects of the administered insulin on glucose levels in blood suggested the presence of insulin resistance. Also, the hyperglycemia and hyperinsulinemia observed in the GK rats at both age groups confirmed the metabolic characteristics of this T2DM model. In previous GK rat studies, skeletal 24,25 and heart 26 muscle as well as liver 27 and adipose tissue 28 were insulin-resistant.
The GK rat exhibits spontaneous moderate hyperglycemia hyperinsulinemia and high plasma triglyceride levels 21,29 without abnormal elevated content of non-esterified fatty acids (NEFA). GK hearts have been consistently reported to be hypertrophied 26,29 , elevated cardiac NEFA and triglycerides with alteration of cardiac structure and function 30 . Thus, there are several evidence supporting that the GK rats represent a reliable model of diabetic cardiomyopathy 29 in absence of obesity. www.nature.com/scientificreports www.nature.com/scientificreports/ The reduced heart weight in diabetic rats was also found in another study on GK hearts 20 with a similar age to our group. These findings are consistent with another rat model of cardiomyopathy in which the reduced mass was attributed to the lack of insulin on heart myocyte growth and protein synthesis 31 . In contrast, GK male rats at 47 weeks were reported to have similar heart weight to the control group 26 . The difference between our group and that of the previous study appears to be age related because the rats of our study were several months younger (28 weeks). Indeed, GK heart weight were reported to be increased, decreased or unaltered for different age groups 29,32 . Mitochondria function. Cardiac mitochondrial dysfunction has been reported in obese subjects with 33 and without insulin resistance 7 , but the contribution of hyperglycemia and obesity to this dysfunction remains to be determined 34 . Also, these reports are relevant for T2DM patients who are not obese and who are regarded to be at high risk for cardiovascular disease 35,36 . An age-dependent relationship between cardiomyopathy and heart mitochondrial dysfunction was reported in cardiomyopathic hamsters at 17 and 30 weeks, but not as early as 4 weeks 15 . Thus, we investigated diabetic heart subpopulations of mitochondria in GK rats, a non-obese model of T2DM, at 18 and 28 weeks. The unchanged mitochondrial function and enzymatic activity of the electron transport chain complexes (Fig. 6) are consistent with data of a study on aorta mitochondria of GK rats showing a respiration rate with complex I and β oxidation substrates similar in GK and control rats 37 .
In both group of rats, SSM and IFM respiration rate were similar with a trend to be higher in IFM. This was mainly related to the CS and SDH activities which were similar in both SSM and IFM ( Table 2). The effect of the specific activity of mitochondrial marker enzymes (i.e. CS and SDH) on mitochondrial respiratory rate was confirmed by the significant correlation between mitochondrial enzyme marker and respiration rate. In a previous study 38 on dog heart mitochondria similar respiratory rates in SSM and IFM were accompanied by similar CS activity of SSM and IFM.
Susceptibility of heart SSM and IFM to cardiomyopathy has been shown to be different in obese and insulin resistant animal models. Cardiomyopathy studies with animals without insulin resistance reported dominant dysfunction of IFM in comparison to SSM in heart of hamster 15 and mouse with T1DM 16,39 . In T1DM mice, mitochondrial respiration with complex I and complex III substrates were reduced only in IFM 16 . In contrast, obese T2DM mice had reduced state 3 respiration rate and electron transport chain activities in SSM with no change in IFM 17 . Also, in heart mitochondria 40 of T2DM patients, only SSM respiration rate in presence of complex I or FA substrates was compromised, whereas IFM was preserved in obese and insulin resistant patients. Overall, these data suggest that mitochondrial function is spatially sequestered in T2DM heart. Differences associated with animal models and factors such as obesity are determinant of the mitochondrial dysfunction. oxidative stress. Diabetic cardiomyopathy is characterized by alteration of substrate utilization and mitochondrial function accompanied by oxidative stress 41 . A previous GK study reported a high susceptibility of heart www.nature.com/scientificreports www.nature.com/scientificreports/ mitochondria to lipid peroxidation in the presence of induced oxidative damage 22 , whereas lipid peroxidation products are used as biomarkers of oxidative stress. It has been suggested that the greater susceptibility of diabetic heart mitochondria to oxidative damage was caused by a limited antioxidant potential related to a lower co-enzyme Q and glutathione contents 22,42 , but mitochondria function was not examined. In our study, a reduced mitochondrial aconitase activity in GK rats (Table 2) confirmed the presence of heart mitochondrial oxidative stress in this animal model of T2DM. The iron sulfur cluster of aconitase is sensitive to oxidation by superoxide 43 which mediates the inactivation of aconitase. Inactivated aconitase is then degraded by Lon protein in the mitochondrial matrix 44 . Thus, the lower aconitase activity is compatible with a higher inactivation rate induced by oxidative stress. Furthermore, a study in 14 week-old GK rats 45 reported reduced aconitase activity in the aorta and kidney whereas it was unchanged in the heart. In our study, reduced aconitase activity was observed at 18 weeks suggesting a time-dependent process leading to oxidative stress of heart mitochondria that is not evident at 14 weeks. As with our observations at 18 weeks, aconitase decreased activity is present in 28 weeks-old rats, too, confirming that in TDM2 non-obese rats ongoing oxidative stress impairs the function of this important enzyme.
The immunoblotting assay on mitochondrial ACO-2 confirmed a similar aconitase protein expression in mitochondria of both diabetic and control rats (Fig. 2). Thus, it is reasonable to exclude the possibility that the reduced aconitase activity in GK was related to low aconitase expression caused by other factors as, for instance, hypoxia inducible factor which might regulate the aconitase expression 46,47 .
Animal models of obesity and T2DM showed an increase of oxidative stress in the presence of cardiac lipid accumulation with increased fatty acid oxidation 6,33,48 . In our study, myocardial mitochondria enzymatic activity of GK rats suggested the presence of oxidative stress even when the animals were consuming normal diet and not obese. An increase of ROS production could be related to insulin resistance rather than to a myocardial overload 41 ; GK rats have been described as myocardial insulin-resistant 19 . In support of this view, in the absence of insulin resistance, a mouse model of T1DM 49 and patients with type 1 diabetes 50 did not exhibit oxidative stress. www.nature.com/scientificreports www.nature.com/scientificreports/ The deleterious interaction of ROS and endoplasmic reticulum stress are suggested to contribute to diastolic dysfunction in diabetic cardiomyopathy 51 , as has been reported in GK rats 20 . Cardiomyopathy observed in GK appears mild but still significantly affects contractility. In the presence of an induced infarction, contractility dysfunction was greater in GK than that in control rats, suggesting that heart failure progression is accelerated in this animal model of T2DM 52 .
An increase of ROS production has been reported to be related to mitochondrial dysfunction due to uncoupling 53 because superoxide can also be produced as a byproduct of oxidative phosphorylation. In our study, oxidative stress is present in the absence of mitochondrial uncoupling (Table 2). Our data support the view that oxidative stress was not generated by dysfunction of both subpopulations of mitochondria. Previous studies reported that diabetes selectively causes oxidative stress in IFM mitochondria, accompanied by a reduction of respiration, transmembrane potential, and an increase of mitochondrial transition pore opening 16,39 . It is noteworthy that mitochondrial transition pore opening has been suggested to be a consequence of oxidative damage rather than the cause 54 .
Therefore, T2DM did not induce an overt mitochondrial impairment in GK rats, suggesting that in lean insulin-resistant individuals mitochondria retain healthy features. Indeed, it has been shown that heart mitochondria of GK rats are even more resistant to calcium overload than are controls 55 , thus indicating that these mitochondria can better counteract the deleterious effects of diabetes.
In presence of obesity and T2DM conditions mitochondrial dysfunction occurs whereas in the absence of obesity, heart mitochondrial function is unchanged and highlights the critical role for obesity in T2DM. Therefore, our study suggests that insulin resistance does not lead to mitochondrial dysfunction in absence of obesity.

Animals.
A non-obese model of type 2 diabetes mellitus (T2DM), Goto-Kakizaki (GK) rats, and Wistar (W) colony rats (Charles River) as control group were used in this study. GK rats' manifest spontaneous skeletal muscle and hepatic insulin resistance, mild hyperglycemia, and normal lipidemia. The genetic background of the GK rats is that of Wistar. The GK rats were obtained by selective breeding of Wistar rats with the highest blood glucose levels during an oral glucose tolerance test over many generations 56 .
The insulin tolerance test was performed on regularly fed rats which were not starved overnight prior to the experiment. Experiment was initiated around 9:00AM, rats were not allowed to eat during the course of the experiment. Rats were challenged with a subcutaneous injection of insulin (1.85 U insulin/Kg of body weight). Blood samples from tail vein puncture were obtained at different time points after insulin injection to measure glucose concentration 57,58 .
Twelve male GK and W rats were housed in the Animal Resource Center facilities of Case Western Reserve University with 12:12-h light-dark cycle and were fed a standard diet chow (Prolab Isopro RMH 3000, LabDiet, St. Louis, MO) ad libitum. GK and W rats were euthanized by decapitation at 18 wk (n = 6) and 28 wk (n = 6) of age. Mitochondrial isolation. The subsarcolemmal (SSM) and interfibrillar mitochondria (IFM) were isolated from heart muscle as described previously 62 with modifications including the use of a modified Chappell-Perry buffer 63 and trypsin to treat myofibrillar pellets 38 . Mitochondrial protein concentration was determined using the Lowry method with bovine serum albumin as standard 64 . oxidative phosphorylation assays. Mitochondrial oxygen consumption was measured using a Clark-type oxygen electrode (YSI model 53) embedded in a glass metabolic chamber containing 0.1-0.25 mg mitochondrial protein in a final volume of 0.5 mL of respiration buffer. The chamber temperature was maintained at 30 °C with a circulating water bath 61 .
The respiration rate of SSM and IFM was measured in the presence of substrates and inhibitors 38 . The assays with complex II, III, and IV substrates were performed with rotenone to inhibit complex I.
Glutamate was used because among the complex I substrates, it is the only substrate that does not require any further additions for the analysis of oxidative substrate. It has its own transporter through the inner membrane and in the matrix it is dehydrogenated by glutamate dehydrogenase yielding NADH, which is then oxidized by complex I of the ETC 65 .
Both low and high ADP concentration were used to stimulate mitochondrial respiration. Because a high ADP concentration does not allow to observe the transition from state 3 to 4 in polarographic systems with O 2 concentrations near air-saturation, a low ADP concentration was selected to reach state 3 in coupled mitochondria avoiding O 2 depletion within the chamber during State 4. Also, with saturated ADP concentration (i.e. High ADP), the state 3 respiratory rate is expected to be higher than that obtained with unsaturated concentration because of the specific ADP affinity to adenine nucleotide translocator and mitochondrial enzymes. Thus, in our study, State 4 is equivalent to a leak state (L T ) with ATP hydrolysis and state 3 indicates the condition under which coupled mitochondria are stimulated with ADP (P) 66 .
Respiratory control ratio (RCR, State 3 divided by State 4) was used to determine the coupling of mitochondrial oxidation and phosphorylation. An enzymatic method 67 was used to determine the concentration of ADP and AMP for the calculation of the ADP/O ratio (number of ADP moles added for the number of moles of oxygen atom consumed), which is an index of the efficiency of oxidative phosphorylation 68 . preparation of samples and enzymatic assays. Mitochondrial enzyme activities were measured as described previously 61,69-71 for both SSM and IFM. Citrate synthase (CS) 69 and electron transport chain enzyme activities were measured in mitochondrial samples treated with cholate: CI, complex I -rotenone-sensitive; CIII, complex III -antimycin A-sensitive decylubiquinol-cytochrome c reductase; NCR, rotenone-sensitive NADH-cytochrome c reductase; SCR, antimycin A-sensitive succinate-cytochrome c reductase; NFR, NADH-ferricyanide reductase; SDH, succinate dehydrogenase; aconitase; CII, thenoyltrifluoroacetone(TTFA)sensitive succinate-Q reductase; CII + Q, TTFA-sensitive complex II with exogenous coenzyme Q 1 . The donors and acceptors span specific regions of the ETC [69][70][71] . The ETC activity components were determined using biochemical kinetics principles: CIV activity was determined as a first-order reaction relationship, whereas the activity of the other ETC components was determined with a zero-order reaction relationship.
immunoblotting. Isolated SSM and IFM mitochondria from 28 weeks old GK or Wistar rats were diluted 1:3 in loading buffer (4% SDS, 20% glycerol, 160 mM dithioerythrol, 125 mM Tris-Cl (pH 6.8), bromophenol blue 0.004%). Samples were then sonicated for 3 min and subsequently heated for 10 minutes at about 76-78 °C for denaturation. Sample's proteins were then separated by electrophoresis on 4-20% Mini-PROTEAN ® TGX ™ precast polyacrylamide gels (Biorad, Hercules, CA, USA), and then transferred on PVDF membranes, which were blocked for 2 hours with 5% milk in Tris buffered saline with 0.1% Tween 20 (TBS-T). Then incubation with primary antibodies followed overnight at 4 °C. The following antibodies were used: anti-Aconitase-2 rabbit polyclonal antibody (Proteintech, dilution 1:1000); anti-COX IV rabbit polyclonal antibody (Invitrogen, dilution 1:1000). Secondary antibodies (goat anti-rabbit peroxidase conjugate 1:2000, Sigma-Aldrich) were incubated for 1 hour at room temperature. Detection of protein signals was achieved by using the ECL Prime chemiluminescence kit (GE Healthcare) and images acquisition using a Fujifilm Luminescent Image Analyzer LAS4000 System (Fujifilm, Tokyo, Japan). Immunoreactive bands were analyzed for densitometry with Image Studio Lite Software (LI-COR, Nebraska, USA). Proteins quantification was expressed as the relative intensity of protein signals normalized to the expression of the housekeeping gene COX IV.

Statistical analysis.
The results are reported as means ± standard deviation. Differences between control and diabetic rats at different ages were evaluated with two-way ANOVA with Bonferroni-Holm correction for multiple comparisons. Differences between control and diabetic rats at different time points of the insulin test were estimated with a two-tailed student t-test. A regression analysis was performed to determine whether a correlation exists between a mitochondrial marker enzyme (i.e. CS and SDH) and respiration rate. Differences were considered statistically significant at p < 0.05.