Enhanced cardiac protein glycosylation (O-GlcNAc) of selected mitochondrial proteins in rats artificially selected for low running capacity

Abstract

O-linked β-N-acetyl glucosamine (O-GlcNAc) is a posttranslational modification consisting of a single N-acetylglucosamine moiety attached by an O-β-glycosidic linkage to serine and threonine residues of both nuclear and cytosolic proteins. Analogous to phosphorylation, the modification is reversible and dynamic, changing in response to stress, nutrients, hormones, and exercise. Aims of this study were to examine differences in O-GlcNAc protein modification in the cardiac tissue of rats artificially selected for low (LCR) or high (HCR) running capacity. Hyperinsulinemic-euglycemic clamps in conscious animals assessed insulin sensitivity while 2-[¹⁴C] deoxyglucose tracked both whole body and tissue-specific glucose disposal. Immunoblots of cardiac muscle examined global O-GlcNAc modification, enzymes that control its regulation (OGT, OGA), and specific proteins involved in mitochondrial oxidative phosphorylation. LCR rats were insulin resistant disposing of 65% less glucose than HCR. Global tissue O-GlcNAc, OGT, OGA, and citrate synthase were similar between groups. Analysis of cardiac proteins revealed enhanced O-GlcNAcylation of mitochondrial Complex I, Complex IV, VDAC, and SERCA in LCR compared with HCR. These results are the first to establish an increase in specific protein O-GlcNAcylation in LCR animals that may contribute to progressive mitochondrial dysfunction and the pathogenesis of insulin resistance observed in the LCR phenotype.

the posttranslational modification of proteins by O-linked β-N-acetyl glucosamine (O-GlcNAc) has emerged as an important regulatory mechanism for numerous cellular processes, including the modulation of critical signaling cascades, stress responses, and metabolism (12, 17). Analogous to phosphorylation, O-GlcNAc is dynamic and responsive to a variety of stimuli including genetic and environmental factors (5, 19, 33). O-GlcNAc modification is capable of affecting protein locale, stability, protein-protein interactions, and most importantly function. Protein O-GlcNAc modification is mediated in part by the activity of the hexosamine biosynthetic pathway (HBP), which is sensitive to the availability of energy substrates such as glucose and free fatty acids (24, 31). Increases in O-GlcNAcylation have been linked to metabolic disease states with perturbed glucose homeostasis such as Type 2 diabetes (7, 25).

Like O-GlcNAcylation, mitochondrial dysfunction has been consistently suggested to underlie the development of Type 2 diabetes. Reduced mitochondrial activity and aberrations in mitochondrial protein regulation have been linked to the development of insulin resistance (1, 34, 40). These mitochondrial impairments are speculated to be the result of abnormal glucose and fatty acid metabolism leading to the increased production of reactive oxygen species (ROS), glucotoxins, depressed ATP synthesis, and reduced insulin action (38). Exposure of cardiac myocytes to hyperglycemic conditions results in increased O-GlcNAc modification of mitochondrial proteins and mitochondrial dysfunction (16, 22). Likewise, mitochondrial dysfunction with aging results in enhanced O-GlcNAcylation in multiple tissues, suggesting an interaction between O-GlcNAcylation and age-related reductions in metabolic function and possibly the development of chronic disease (15). Indeed, enhanced protein O-GlcNAcylation is now implicated in ischemia, vascular dysfunction, and heart failure (9, 52, 55).

As an increase in protein O-GlcNAcylation (46) and mitochondrial dysfunction (43) have been independently linked to the development of insulin resistance, the aim of this study was to evaluate the connection between protein O-GlcNAc and insulin resistance. This was achieved by examining O-GlcNAcylated protein levels in rats artificially selected for either low (LCR) or high running capacity (HCR). LCR and HCR rats provide a contrasting genetic model system for assessing the impact of innate aerobic exercise capacity on a wide spectrum of disease risk factors (53). Previous research has shown that low aerobic fitness is a strong predictor of metabolic disease (30, 47). More specifically, LCR rats are predisposed to becoming obese and developing insulin resistance and cardiovascular dysfunction (37, 53). Functional cardiovascular reductions in LCR animals may be related to an increase in O-GlcNAc (6, 29). In this study, it is hypothesized that enhanced O-GlcNAcylation of specific mitochondrial proteins may lead to the impairment of insulin sensitivity and contribute to metabolic derangements observed in LCR.

MATERIALS AND METHODS

Animals.

All procedures were approved by the University of Calgary Animal Care and Use Subcommittee and followed Canadian Council Animal Care guidelines for the care and use of laboratory animals. The selection of rats yielding HCR and LCR has been previously described in detail (28). In brief, selection for LCR and HCR was based upon distance run to exhaustion using a velocity ramped exercise test. Animals were phenotyped at 11 wk of age for maximal running capacity on a motorized treadmill beginning at 10 m/min and increasing 1 m/min every 2 min until exhaustion, where slope was kept constant at 15°. We chose 13 of the lowest and 13 of the highest capacity rats of each sex from the founder population, and two-way (divergent) selection was applied to generate each model. After 11 generations of selection, each group differed in running capacity by 347% (53). For the present experiments, male rats (n = 7–10 per group) from generation 19 were obtained. All tests for intrinsic running capacity were performed at the University of Michigan (Ann Arbor, MI) in accordance with the University of Michigan Committee guidelines on the use and care of animals. Rats were then shipped to the University of Calgary for further analyses. Upon arrival at the University of Calgary, animals were housed individually and maintained at 22°C on a 6:00 AM to 6:00 PM light cycle without access to a running wheel. Animals were given standard chow and water ad libitum and were killed at 50 wk of age.

Surgical procedures.

To assess insulin sensitivity, we surgically implanted indwelling catheters to conduct hyperinsulinemic-euglycemic (insulin) clamps. Animal surgeries were conducted 5 days before the experimental protocol. Surgical procedures were performed as described previously for arterial and venous catheterizations (23, 39). In brief, animals were anesthetized with a 50:5:1 (vol/vol) mixture of ketamine, xylazine, and acepromazine, and the left common carotid artery and right jugular vein were catheterized with PE50 tubing. Catheters were exteriorized and secured at the back of the neck, filled with heparinized saline (150 U/ml), and sealed with a stainless steel plug. Immediately postsurgery, each animal received 75 mg/kg sc ampicillin to prevent infection. After surgery, animal weights and food intake were monitored daily, and only the animals in which presurgery weight was restored were used for experiments. A schematic of the animal procedures are shown in Fig. 1.

Fig. 1.

A: experimental schematic of the hyperinsulinemic-euglycemic clamp procedure. Animals were fully recovered and conscious at the time of study. Insulin, glucose (D50), and glucose tracer 2[¹⁴C]deoxyglucose (2[¹⁴C]DG) were infused into the jugular vein while blood was sampled from the carotid artery. Black circle indicates a Y-connector. B: time course of infusion and final tissue measurements. Euglycemia was maintained at 6 mM at the time of tracer infusion. OGT, O-linked N-acetylglucosamine transferase; OGA, O-GlcNAcase.

Dual energy X-ray absorptiometry.

Body composition of recovered rats, including body fat and lean mass, was determined 18 h before start of experiments by dual energy X-ray absorptiometry (DEXA). Rats were lightly anesthetized with isoflurane for the duration of the scan, lasting ∼2 min, and body composition was measured with software for small animal analysis (Hologic QDR 4500; Hologic, Bedford, MA). Total lean and fat mass are reported as percentages.

Hyperinsulinemic-euglycemic clamp.

A schematic depicting the clamp procedure is shown in Fig. 1, A and B. This procedure has been described previously in detail (23). In brief, on the day of the study rats were fasted for 10 h and given free access to water. Approximately 1 h before the experiment, catheters were flushed with heparinized saline (20 U/ml) and connected to polyethylene (PE50) and silastic tubing for infusions and sampling. Rats were then placed back in the cage until the commencement of the experimental protocol. Throughout the experimental protocol, rats were conscious and unrestrained. A baseline arterial blood sample (100 μl) was obtained for the measurement of plasma glucose, and nonesterified fatty acids (NEFA) and hematocrit. The insulin clamp was initiated with a constant infusion of insulin at 4 mU·kg⁻¹·min⁻¹ maintained throughout the experiment (Humulin R; Eli Lilly, Toronto, Canada). Stable euglycemia was maintained during each insulin clamp by measuring blood glucose. At t = 0 min, blood was sampled (50 μl) to determine arterial plasma glucose and NEFA. At t = 5 min, 2-[¹⁴C]deoxyglucose (2[¹⁴C]DG) was administered as a bolus into the jugular vein to provide an index of tissue-specific glucose uptake. Additional arterial blood samples were obtained from the carotid artery at 7, 10, 15, 20, and 30 min for the measurement of 2[¹⁴C]DG isotope, plasma glucose, and NEFA. The relatively short duration on the clamp procedure would not be expected to increase O-GlcNacylation (2). To prevent declines in hematocrit during the procedure, the erythrocytes taken before isotopic analog infusion were washed in saline and re-infused shortly after each sample was taken. After the last blood sample, rats were anesthetized with pentobarbital sodium, and the heart was rapidly excised, rinsed in saline to remove excess blood, freeze-clamped in liquid nitrogen, and kept frozen at −80°C until further analysis. To examine muscle mass between the two models, individual skeletal muscles were carefully excised, and muscle tissues were weighed twice for accuracy using a calibrated analytical balance (Mettler Instruments, Zurich, Switzerland). Animals with declines in hematocrit >10% were excluded from the study.

Plasma measures.

NEFAs were measured spectrophotometrically according to manufacturer's instructions (Wako NEFA C kit; Wako Chemicals, Richmond, VA). Plasma glucose concentrations were measured by the glucose oxidase method with the use of an automated glucose analyzer (Beckman Instruments, Fullerton, CA).

Tracer analysis.

Hearts were homogenized in 2 ml of 0.5% perchloric acid and centrifuged for 20 min. Supernatants (1.5 ml) were then neutralized using 5 M KOH, and radioactivity in a 250 μl sample was determined by liquid scintillation counting using a Packard Tri-Carb 2900TR Liquid Scintillation Analyzer (PerkinElmer, Boston, MA) with Ultima Gold scintillant (PerkinElmer).

Whole body and tissue-specific substrate kinetics.

The metabolic index of glucose (R_g) was calculated and expressed as described previously (44, 45, 54). Briefly, glucose clearance (K_g) and metabolic (R_g) indexes were calculated from the accumulation of 2[¹⁴C]-deoxyglucose-b-phosphate in cardiac muscle and the integral of the plasma 2[¹⁴C]DG concentration after a 2[¹⁴C]DG bolus. The relationships are defined as follows:

$$\begin{array}{c}Kg={\left[\left[14C\right]DGP\right]}_m\left(t\right)/{\displaystyle \underset{o}{\overset{t}{\int }}{\left[\left[14C\right]DG\right]}_{p}dt}\\ Rg=Kg\times {\left[G\right]}_p\end{array}$$

The subscripts p and m refer to mean arterial plasma and total muscle concentration during the tracer administration period, and [G] is the glucose concentration. R_g for cardiac and peripheral tissues was expressed relative to R_g of the brain, which represents the constant reservoir of glucose uptake under various physiological conditions (54).

Global and mitochondrial protein O-GlcNAc modification.

Heart tissue (∼200 mg) was homogenized in buffer (20 mM NaCl, 20 mM Tris·HCl, 0.1 mM EDTA, 1% Triton X-100, 0.5% sodium deoxycholate, and 0.1% β-mercaptoethanol, pH 7.4) in the presence of SIGMAFAST Protease Inhibitor Cocktail (Sigma Aldrich, St. Louis, MO), HALT Phosphatase Inhibitor Cocktail (Fisher Scientific, Ottawa, Canada), and O-GlcNAcase (PUGNAC) Inhibitor (Toronto Research Chemicals, Toronto, Canada). After a centrifugation at 14,000 rpm (10 min, 4°C), supernatant protein concentrations were determined by the Bradford method (Bio-Rad, Rockford, IL). Equal quantities of protein (75–100 μg) were loaded into precast 3–8% Tris-Acetate gels (Invitrogen, Carlsbad, CA) and electrophoresed for 1 h at 200 V. Proteins were then electrotransferred for 2 h at 30 V onto nitrocellulose membranes and probed overnight with primary antibodies. For examining global O-GlcNAc expression, we used a monoclonal antibody directed against RL2, anti-O-GlcNAc (Abcam, Cambridge, MA), at a 1:1,000 dilution. For examining O-GlcNAc regulation, we used primary antibodies directed against O-linked N-acetylglucosamine transferase (OGT) (Sigma Aldrich) and O-GlcNAcase (OGA) (Santa Cruz Biotechnology, Santa Cruz, CA) at a dilution of 1:1,000 each. Equal loading was verified with Ponceau red stain and by detection of loading control protein α-actin (Sigma Aldrich) at a dilution of 1:2,000. Membranes were washed in 0.05% Tween-PBS buffer and incubated with secondary antibodies conjugated with horseradish peroxidase. For wheat germ agglutinin (WGA) studies, WGA-conjugated beads (Sigma Aldrich) were used to pull down O-GlcNAcylated proteins. Whole heart lysate (500 μg protein) was incubated overnight while rotating at 4°C with the WGA beads. After incubation, the beads were washed four times, and the isolated O-GlcNAc modified proteins were analyzed by Western blot as described above. For examining WGA isolated proteins as well as total protein levels, antibodies directed against Complex I (NDUFA9) (Invitrogen) and Complex IV (Subunit I) (Mitosciences, Eugene, OR) were used at a dilution of 1:1,000. Anti-voltage-dependent anion channel (VDAC) (Abcam) and anti-sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA) (Santa Cruz Biotechnology) antibodies were used at a dilution of 1:1,000 and 1:2,000, respectively. All signals were detected using enhanced chemiluminescence substrate (ThermoScientific, Rockford, IL). Chemiluminescence was digitally captured using the Chemigenius² BioImaging System (Syngene, Frederick, MD) and band densitometry was measured using the Gene Tools software (Syngene).

The selection of specific O-GlcNAc modified proteins was based on distinct criteria. Proteins either appeared in dbOGAP, Database of O-GlcNACylated Proteins and Sites (http://cbsb.lombardi.georgetown.edu/OGAP.html) or were previously studied, quantified, and reported in the literature. Selected proteins were also relevant to the metabolic function of the heart.

Citrate synthase assay.

Citrate synthase (CS) activity was used as a proxy for mitochondrial content. Flash-frozen heart tissue from HCR and LCR animals was used in measurement of CS enzyme activity. All methodology was carried out as previously described (49) at 25°C with CS activity expressed as micromoles per minute per gram wet mass.

Statistical analysis.

Values represent means ± SE. Comparisons between groups were performed by a one-way ANOVA. Differences within the ANOVA were determined by Tukey's post hoc test. Differences were considered significant if P < 0.05. All data was examined for normality. In cases where data were not normally distributed, significant differences between groups were assessed by an ANOVA on ranks with a Dunn's post hoc test.

RESULTS

Exercise capacity and body composition.

Animal characteristics are reported in Table 1. Rats from generation 19 were examined at 11 and 50 wk of age. At 11 wk of age, the body mass LCR were considerably greater compared with HCR(318 ± 9 vs. 252 ± 9 g). During a standardized run to exhaustion, HCR animals covered a distance of 1,436 ± 16 m compared with LCR whose mean distance was 367 ± 11 m (28). In addition, running speed and running time were superior in HCR animals (P < 0.05). When examined at 50 wk of age, LCR animals gained on average 50 g more than their HCR counterparts (468 ± 20 vs. 356 ± 16 g). Analysis of body composition by DEXA confirmed LCR animals had a greater fat mass (21 vs. 13%) compared with HCR. Lean mass was higher in LCR in all tissues examined including the heart, soleus, gastrocnemius, tibialis anterior, and extensor digitorum longus. Muscle mass in LCR rats was approximately ∼25% greater than HCR rats and did not depend upon the muscle examined (Table 1).

Table 1.

Characteristics of male LCR and HCR rats

Phenotypic Data	LCR	HCR
11 wk of Age
Generation	19	19
Body mass, g	318.5 ± 9.4	252.2 ± 8.8*
Best running time, min	23.8 ± 0.5	59.1 ± 0.4*
Best running distance, m	367.1 ± 11.4	1436.0 ± 16.5*
Best running speed, m/min	21.3 ± 0.2	39.1 ± 0.3*
50 wk of Age
Body Mass, g	468 ± 20	356 ± 16*
% Body Fat	21.4 ± 2.0	13.5 ± 1.5*
Heart mass, mg	1249 ± 4	1088 ± 5*
Soleus mass, mg	169 ± 6	139 ± 10*
Gastrocnemius mass, mg	2,236 ± 81	1,645 ± 98*
Tibialis anterior mass, mg	802 ± 27	648 ± 49*
Extensor digitorum longus mass, mg	218 ± 11	162 ± 15*
Fasting glucose, mM	6.6 ± 0.3	6.4 ± 0.4
Fasting NEFA, mM	0.58 ± 0.05	0.70 ± 0.06

Data represent means ± SE. NEFA, nonesterified fatty acids.

^*P < 0.05 for low capacity running (LCR) vs. high capacity running (HCR) (n = 7–10 animals per measure).

Impaired insulin sensitivity in LCR.

Following a 10 h fast, arterial blood glucose and fasting NEFA were comparable between LCR and HCR. To examine insulin action in LCR and HCR animals, insulin clamps were performed. During the last 30 min of the insulin clamp, glucose levels were held constant in LCR and HCR (Fig. 2A). Analysis of NEFA indicates no differences between HCR and LCR at individual time points during the clamp. However, collective measures (0–30 min) show a main effect of lower NEFAs in the HCR group (P < 0.05), suggesting a slightly greater fatty acid disposal in these animals (Fig. 2B). Analysis of glucose infusion rates (GIR) showed considerable insulin resistance in LCR, with animals disposing of 65% less glucose compared with HCR (P < 0.05) (Fig. 2C). Metabolic indexes of tissue-specific glucose utilization showed lower rates of glucose utilization in the soleus (P < 0.05) (Fig. 3A), a muscle composed of 89% type I, 11% type IIa, and 0% type IIb fibers in the rodent (4). Other tissues showed a trend; however, there were no significant differences in glucose disposal between HCR and LCR groups (Figs. 3, B and C).

Fig. 2.

A: arterial blood glucose concentration during the hyperinsulinemic-euglycemic clamp of low running capacity (LCR) and high running capacity (HCR) rats. B: arterial plasma concentration of nonesterified fatty acids (NEFA). Arterial measures were obtained during the last 30 min of the hyperinsulinemic-euglycemic clamp; no differences were apparent between groups for either measurement. C: glucose infusion rate (GIR) required to maintain euglycemia. This measurement directly assesses whole body glucose utilization in response to a set insulin dose. Data represent means ± SE for 5–7 animals/group. *P < 0.05 between LCR and HCR.

Fig. 3.

Metabolic index of insulin-stimulated glucose utilization into skeletal muscles. R_g, an index of muscle glucose uptake, was measured in conscious, freely moving rats. A: soleus, B: vastus lateralis (vastus), C: gastrocnemius (gastroc) and D: heart in LCR and HCR animals. Data are expressed relative to the values obtained for HCR rats and are means ± SE for 5–7 animals/group.

Global and mitochondrial protein O-GlcNAc modification.

Global O-GlcNAcylation of proteins in total heart homogenate was evaluated with the anti-O-GlcNAc antibody RL2 to compare LCR and HCR groups as previously reported (5). Numerous bands were observed between ∼50 and 200 kDa molecular mass (Fig. 4A). Graphical data are shown normalizing values to the loading control α-actin, which did not change with phenotype (P > 0.05). When LCR and HCR groups are compared, global O-GlcNAcylated protein levels are not different (P > 0.05) (Fig. 4B). Levels of the enzymes that control the regulation of O-GlcNAc were also quantified. OGT catalyzes the transfer of O-GlcNAc to substrate proteins, and conversely, OGA catalyzes removal of the moiety. No observable differences in OGT and OGA expression were found between LCR and HCR groups (Fig. 4C). Incubation of heart tissue homogenate, from LCR and HCR, with WGA-conjugated agarose beads allowed for the isolation of O-GlcNAc-modified proteins. WGA beads bind to O-GlcNAcylated proteins due to its high affinity of WGA for the terminal portion of O-GlcNAc residues. This technique is commonly used to pull down O-GlcNAcylated proteins (41). As shown in Fig. 5, Western blots were then used to examine the proportion of O-GlcNAc modified Complex I (NDUFA9), Complex IV (Subunit I), VDAC, and SERCA compared with respective total protein levels in the same tissue homogenate. As demonstrated in Fig. 5, A–D, levels of protein-specific O-GlcNAc modification are increased in the LCR model (P < 0.05) compared with the HCR model.

Fig. 4.

A: sample immunoblot of global O-GlcNAc modified proteins between ∼50 and 200 kDa molecular mass between HCR and LCR. B: graphical representation of global O-GlcNAc levels in cardiac tissue as normalized to α-actin, C: levels of OGT: O-linked N-acetylglucosamine transferase and (OGA) O-GlcNAcase; responsible for the addition and removal of O-GlcNAc modification. Values are normalized to loading control, α-Actin. Values represent means ± SE and n = 7 animals per measure.

Fig. 5.

O-GlcNAcylation of mitochondrial proteins is increased in LCR hearts as determined by Wheat-Germ Agglutinin pull-down. A: Complex I (NDUFA9), B: Complex IV (Subunit I), C: VDAC and D: SERCA are accompanied with representative immunoblots. All values are normalized to total protein levels; α-Actin was also run as a secondary control. Values represent means ± SE. *P < 0.05 for LCR vs. HCR and n = 6–7 animals per measure.

CS activity.

To ascertain similar mitochondrial protein content between LCR and HCR hearts, we assessed CS enzyme activity. No significant variation in the activity of CS was found between cardiac preparations from LCR and HCR rats. CS activity was 61.19 ± 4.50 μmol·min⁻¹·g⁻¹ in the LCR group and 73.57 ± 4.69 μmol·min⁻¹·g⁻¹ in the HCR group (P > 0.05). No difference in CS activity between groups suggests there is similar mitochondrial content in each phenotype.

DISCUSSION

Artificial selection of rats for divergent running capacity results in strikingly distinct phenotypes that go well beyond aerobic capacity per se. The HCR phenotype displays superior health, higher aerobic capacity, resistance to environmental challenges, enhanced insulin sensitivity, and protection from oxidative damage (29, 37, 48, 49). In contrast, the LCR phenotype displays characteristics indicative of numerous metabolic disease states including hyperlipidemia, impaired glucose tolerance, hypertension, and elevated oxidative stress (11, 32, 48, 49). Although differences have been demonstrated between these selection groups in numerous tissues at both the transcriptome (27) and proteome levels (8) that likely contribute to the superior health of HCR rats, alterations at the posttranslational level have not been investigated. As such, we assessed O-GlcNAcylation in cardiac muscle of LCR and HCR rats. Novel findings show no change in global protein O-GlcNAcylation between selection groups, yet profound differences in the modification of key mitochondrial proteins in cardiac muscle. Specifically, LCR rats have increased levels of O-GlcNAcylation on numerous mitochondrial proteins compared with HCR, supporting the view that the posttranslational modification is closely intertwined with metabolism and possibly the development of insulin resistance.

Recent research shows that insulin resistance causes an increase in cardiac O-GlcNAcylation that can be reversed through exercise despite persistent hyperglycemia (6). As such, we investigated global, protein-specific O-GlcNAcylation as well as levels of key regulatory enzymes involved in protein modification. Global protein glycosylation was comparable between LCR and HCR animals, as were protein levels of OGA and OGT. OGT is a highly conserved protein catalyzing the addition of O-GlcNAcylation, while OGA is the sole protein responsible for its removal (56). These findings suggest that protein O-GlcNAcylation is predominantly substrate driven in LCR and therefore presumably determined by the magnitude of HBP flux, rather than by changes in OGT and/or OGA. In the last several years, numerous studies have associated protein O-GlcNAcylation with metabolic syndrome and insulin resistance (18, 46, 50). At 11 wk of age, LCR animals exhibited higher body adiposity and lower running capacity compared with their HCR counterparts. Without any alterations to diet or exercise levels between groups, differences between LCR and HCR continued to diverge with age. Insulin clamps at 50 wk in combination with a radiolabeled glucose tracer provide further evidence of whole body and tissue-specific insulin resistance in LCR animals. Although these animals have been extensively studied in the past, this is the first time they have undergone an insulin clamp procedure. Assessment of in vivo insulin action at high physiological insulin levels reveals LCR rats dispose 65% less glucose compared with HCR. A drop in the GIR of >50% under such conditions is indicative of systemic insulin resistance (44). Examination of individual tissues reveals a reduction of insulin-stimulated glucose uptake in slow-twitch skeletal muscle but not the heart. This finding replicates that of Schwarzer et al. (42), who also concluded that LCR animals lack cardiac insulin resistance despite systemic insulin resistance.

It is widely believed that O-GlcNAcytion is substrate driven (35); yet we observed no differences in basal glucose levels or cardiac-specific insulin resistance. This raises two alternative possibilities. First, it may be that O-GlcNAcylation precedes insulin resistance in this tissue. Evidence supporting the notion that O-GlcNAcytion causes rather than results from insulin resistance is garnered from experiments where pharmacological elevation of O-GlcNAc levels results in muscle insulin resistance (3). Second, although there were no differences in extracelluar glucose between HCR and LCR, previous work demonstrates profound differences in enzyme activity, capillary density, and intracellular metabolism between these selection groups (20). These differences are likely contributing to enhanced flux through the HBP in LCR. Upon entering cells, a small proportion of glucose (∼2–6%) is funneled through HBP. The first two steps involved in HBP generation are shared with glycolysis, diverging at fructose-6-phosphate, which is eventually converted to uridine 5′-diphosphate (UDP)-GlcNAc, the donor molecule for OGT. In this manner, the regulation of O-GlcNAcylation is sensitive and dependent on the concentration of glucose entering the cell. Slight increases in UDP-GlcNAc cause O-GlcNAcylation of a myriad of proteins. Notably, hyperglycemia can lead to enhanced oxidative stress, specifically mitochondrial superoxide and ROS production. Mitochondrial superoxide production inhibits glyceraldehyde-3-phosphate dehydrogenase, a key enzyme in the glycolytic pathway, which further contributes to increased HBP flux by attenuating glycolysis (13). Previous studies have shown HCR animals have increased levels of skeletal muscle superoxide dismutase that likely reduces DNA oxidative damage in these animals (49). Combined, these data support the notion that insulin resistance in LCR may drive increases in HBP, UDP-GlcNAc synthesis, and protein modification in the heart.

Key findings of the present study show targeted modification of mitochondrial proteins in LCR. Data support the hypothesis that metabolic dysregulation and subsequent insulin resistance is occurring in LCR animals, in part, due to the O-GlcNAcylation of key mitochondrial proteins. Analysis of Complexes I and IV, VDAC, and SERCA showed greater levels O-GlcNAcylation in LCR compared with HCR. These data are supported by a recent study, which found altered patterns of O-GlcNAcylation of mitochondrial proteins after treatment of myoblasts with high glucose (16). Additionally, Hu et al. (22) report increases in O-GlcNAc modification of Complex I (NDUFA9) and Complex IV after high glucose exposure and that they were accompanied by reduced mitochondrial Complex I function and lower cellular ATP levels. This suggests that increases in O-GlcNAc modification of Complexes I and IV have a significant effect on functional enzyme activity and likely impair oxidative phosphorylation. These in vitro results, together with previous data, suggest compromised mitochondrial function in LCR animals (20, 51, 53). Given the considerable increase in mitochondrial O-GlcNAcylation in the present study, CS activity was examined as a surrogate marker of mitochondrial content. We found no detectable difference between LCR and HCR groups. These data are supported by previous research showing cardiac CS activity to be similar in between LCR and HCR rats (37).

The nature and purpose of O-GlcNAcylation in the heart remain controversial. Specifically we question whether the modification is protective or detrimental to cardiac function. Current evidence indicates that O-GlcNAc levels are increased as a result of stress, lending support to the view that the modification may be protective (36, 58). In LCR animals, enhanced body adiposity and insulin resistance imply a state of low-grade inflammation, which could promote O-GlcNAcylation. O-GlcNAc modifications are thought to be protective to the heart in the short term by limiting substrate flux, decreasing mitochondrial function, and reducing excessive ROS production. This premise is supported by recent work on VDAC, showing it to be a targeted site for O-GlcNAcylation (26). It has been previously reported that O-GlcNAc modified VDAC may be serving to maintain mitochondrial integrity (10), possibly to increase mitochondrial stability and ROS protection in the presence of metabolic malfunction. Although the O-GlcNAc modification may be protective for acute stress, continued protein modification is thought to precipitate systemic metabolic abnormalities (18, 57).

Along with metabolic abnormalities, LCR animals exhibit a variety of intrinsic functional abnormalities compared with their HCR counterparts. A recent study by Koch and colleagues (29) shows that LCR animals have a decreased contractile function and impaired Ca²⁺ signaling and handling. Evidence of increased SERCA O-GlcNAcylation in LCR animals may play a role in the development of cardiomyocyte dysfunction as a result of the metabolic derangements and abnormal glucose use in these animals. Hu et al. (21) found that adenoviral transfer of OGA (i.e., the enzyme that removes O-GlcNAc modification) reversed the excessive O-GlcNAc modification associated with diabetes, blunted contractile abnormalities, and improved intracellular calcium handling in the diabetic myocardium. To address the molecular mechanism, they found that SERCA levels were augmented by adenovirus-mediated overexpression of OGA. Thus, calcium handling might be normalized by restoration of SERCA expression. More recent evidence confirms that SERCA is O-GlcNAc modified in the heart, yet the impact of this modification on protein function and activity has yet to be elucidated (14). We hypothesize that the functional changes shown by Koch et al. (29) in LCR animals may be related to impaired function of SERCA and its O-GlcNAcylation. This provides a possible conceptual framework for future investigation into the functional role of O-GlcNAc in the heart.

This study raises a number of intriguing questions and future directions regarding genetics, exercise capacity, insulin resistance, and O-GlcNAcylation. Despite no regular exercise training, it appears that an enhanced innate aerobic capacity in HCR animals results in reduced O-GlcNAcylation of key mitochondrial proteins. In contrast, LCR animals have elevated O-GlcNAcylation that would be expected to exacerbate mitochondrial dysfunction and impair cardiac glucose uptake, ultimately resulting in declines in cardiac function. This could lead to a perpetual reduction in aerobic capacity and muscle fatigue over time, thus fueling a vicious cycle of inactivity leading to insulin resistance, mitochondrial dysfunction, muscle weakness, fatigue, and a lack physical activity. These data could explain the substantial divergence in traits between models. From a clinical perspective, future studies examining the ability of exercise training to reverse aberrant O-GlcNAcylation in insulin sensitive tissues is warranted. In conclusion, our results are the first to establish an increase in specific protein O-GlcNAcylation in LCR animals and are consistent with the notion that such posttranslational modifications represent an underlying mechanism in the development of metabolic disease susceptibility in the LCR phenotype. Uncovering approaches for manipulation of protein O-GlcNAcylation may reveal new diagnostic tools and therapeutic targets for insulin resistance.

GRANTS

This study was supported by the Canadian Institutes of Health Research (MOP 79397) and the Alberta Diabetes Association. J. Shearer holds salary support awards from the Alberta Heritage Foundation for Medical Research, Heart and Stroke Foundation of Canada, and the Canadian Diabetes Association. The LCR-HCR rat model system was funded by the National Institutes of Health (NIH) Grant R24 RR-017718 and is currently supported by NIH Office of Research Infrastructure Programs/OD Grant ROD012098A (to L. G. Koch and S. L. Britton). S. L. Britton was also supported by NIH Grant RO1 DK-077200. The LCR and HCR model can be made available for collaborative study (contact: brittons@umich.edu or lgkoch@umich.edu).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: V.L.J., D.D.B., C.C.H., D.S.H., and J.S. performed experiments; V.L.J., C.C.H., and J.S. analyzed data; V.L.J., D.D.B., C.C.H., D.S.H., and J.S. interpreted results of experiments; V.L.J. and J.S. prepared figures; V.L.J., C.C.H., and J.S. drafted manuscript; V.L.J., D.D.B., C.C.H., D.S.H., R.T.H., L.G.K., S.L.B., and J.S. edited and revised manuscript; V.L.J. and J.S. approved final version of manuscript; D.D.B. and J.S. conception and design of research.