Exercise Protects against Diet-Induced Insulin Resistance through Downregulation of Protein Kinase Cβ in Mice

Physical exercise is an important and effective therapy for diabetes. However, its underlying mechanism is not fully understood. Protein kinase Cβ (PKCβ) has been suggested to be involved in the pathogenesis of obesity and insulin resistance, but the role of PKCβ in exercise-induced improvements in insulin resistance is completely unknown. In this study, we evaluated the involvement of PKCβ in exercise-attenuated insulin resistance in high-fat diet (HFD)-fed mice. PKCβ-/- and wild-type mice were fed a HFD with or without exercise training. PKC protein expression, body and tissue weight change, glucose and insulin tolerance, metabolic rate, mitochondria size and number, adipose inflammation, and AKT activation were determined to evaluate insulin sensitivity and metabolic changes after intervention. PKCβ expression decreased in both skeletal muscle and liver tissue after exercise. Exercise and PKCβ deficiency can alleviate HFD-induced insulin resistance, as evidenced by improved insulin tolerance. In addition, fat accumulation and mitochondrial dysfunction induced by HFD were also ameliorated by both exercise and PKCβ deficiency. On the other hand, exercise had little effect on PKCβ-/- mice. Further, our data indicated improved activation of AKT, the downstream signal molecule of insulin, in skeletal muscle and liver of exercised mice, whereas PKCβ deficiency blunted the difference between sedentary and exercised mice. These results suggest that downregulation of PKCβ contributes to exercise-induced improvement of insulin resistance in HFD-fed mice.


Introduction
Diabetes mellitus, especially type 2 diabetes, is one of the most common chronic diseases worldwide [1]. Diabetes is growing worldwide both in number and significance, due to an increase in economic development and urbanization. Diabetes was reported to affect 366 million people globally in 2011, and this number is expected to rise to 552 million by 2030 in both developed and developing countries [2].
Protein kinase C (PKC) is a family of protein kinases that phosphorylates other proteins at serine and threonine residues [3,4]. PKC family proteins are involved in multiple cellular processes, including metabolism, differentiation, and cell growth. They are classified into subfamilies, including conventional isoforms (a, b, and c) that are dependent on both Ca 2+ and diacylglycerol (DAG) for stimulation, novel isoforms (d, e, h, and g) that are dependent on DAG only, and atypical isoforms (f and i/l) that are independent of Ca 2+ and DAG [3,4]. Abnormal expression of PKC family proteins has been observed in skeletal muscles of patients and animals with diabetes [5,6,7]. Among these PKC isoforms, PKCb protein content was significantly higher, whereas PKCh and PKCg were significantly lower, in muscle of obese patients compared with muscle of lean control subjects, without a corresponding change in membrane-associated PKC activity [5]. The PKCb isoform inhibitor ruboxistaurin, which is the most studied PKC inhibitor, has shown some positive effects on diabetes and diabetic complications in clinical trials [8,9,10]. Our previous studies demonstrated that PKCb deficiency alleviated insulin resistance and obesity in mice [11,12]. Although PKCb is important in both obesity and insulin resistance, its role in exercise-related changes in HFD-induced metabolic disorders has not yet been reported.
Numerous studies have shown that a high-fat diet (HFD) and sedentary behavior increase the risk of obesity and insulin resistance, whereas increased physical activity reduces this risk [13,14,15,16]. The potential mechanism by which exercise attenuates HFD-induced insulin resistance involves increasing insulin sensitivity and glucose transport into contracting skeletal muscles [17]. However, the underlying molecular mechanisms are not fully understood due to the complicated processes involved in exercise [17]. Given the important regulatory role of PKCb in insulin resistance, we postulated that it might also play a role in exercise-induced improvement of insulin resistance. We thereby used PKCb knockout mice and a diet-induced obesity model to test this hypothesis. To our knowledge, this is the first study demonstrating the role of PKCb in exercise-attenuated insulin resistance by using PKCb deficiency mice.

Animals and diet
Production of PKCb -/mice in C57BL/6J background and genotypic determination were performed as described previously [18]. At the age of four weeks, PKCb -/and wild-type (WT) C57BL/6J mice were fed a high-fat diet (HFD) containing 42% of calories from fat (TD88137, Harlan, Madison, WI). At the age of 12 weeks, both WT and PKCb -/mice were randomly assigned into sedentary (SED) or exercise (EX) group for 8 weeks ( Figure  S1). All mice were allowed to eat and drink ad libitum throughout the duration of the study. The mice were housed on a 12:12-hour light-dark cycle in a temperature and humidity controlled vivarium. National Institutes of Health guidelines for the care and use of laboratory animals were strictly followed, and all experiments were approved by the Animal Care and Use Committee at The Ohio State University.

Exercise intervention
Exercise intervention was performed as described previously [19]. Briefly, mice in exercise group were exercise-trained on a motorized treadmill (Columbus Instruments, Columbus, OH) at a speed of 15 m/min, 40 min/day, and 5 days/week for 8 weeks. Mice in SED group were put on the same treadmill without running for 40 min/day and 5 days/week for 8 weeks.
Body weight, tissue weight, food intake and water intake Body weight, food intake, and water intake were recorded weekly during the exercise intervention. 24 hours after the end of the exercise intervention, all mice were fasted overnight and euthanized by CO 2 inhalation overdose. Blood samples were obtained and plasma was collected and stored at 280uC immediately. Heart, liver, calf muscles (gastrocnemius and soleus), thigh muscle (quadriceps femoris and adductor magnus), epididymal fat, inguinal fat, together with brown adipose tissue from the interscapular depot were carefully excised. All the tissue samples were weighed and then immediately frozen in liquid nitrogen.

Magnetic resonance imaging (MRI)
Body fat mass (abdominal cavity) was evaluated by in vivo MRI, as described previously [19]. Briefly, 11.7 T small bore vertical NMR system (BioSpec, Bruker, Ettlingen, Germany) was used. First, mouse was anesthetized with isoflurane (1.5-2.0%) and placed in a 30-mm birdcage coil. After the mouse was positioned in the scanner, a coronal spin-echo localizing sequence was used to identify both kidneys. Finally, from the superior pole of the uppermost kidney to the caudal aspect of the mouse, thirty contiguous 1-mm thick axial slices were obtained using a spin-echo sequence with a 2566256 pixel size (30630 mm). Data were analyzed by National Institutes of Health Image J software.

Glucose tolerance test (GTT) and insulin tolerance test (ITT)
After 8 weeks exercise intervention, a glucose tolerance test (overnight fasting) and insulin tolerance test (6 hours fasting) were performed on all mice as previously described [20]. Briefly, mice were weighed and then injected intraperitoneally with either glucose (2 mg/kg body weight) or insulin (0.5 U/kg body weight). Blood samples were collected through the tail vein and glucose concentrations were measured before and 30, 60, 90 and 120 min after the challenge on an Elite Glucometer (Bayer, Leverkusen, Germany). Area under the curve was calculated using GraphPad software.

Plasma insulin, leptin, adiponectin level and insulin resistance assessment
After overnight fasting, blood samples were collected into EDTA-coated tubes and plasma was collected after centrifugation at 20006g for 15 min. Plasma insulin level was determined following a standard protocol of an ultrasensitive Mouse Insulin ELISA kit (Crystal Chem, Downers Grove, IL) [21]. Leptin level was determined following a standard protocol of the Quantikine Mouse Leptin ELISA kit (R&D, Minneapolis, MN). Adiponectin level was measured according to the manufacturer's instructions using the Quantikine Mouse Adiponectin/Acrp30 ELISA kit (R&D, Minneapolis, MN). Insulin resistance (IR) was calculated using the homeostasis model assessment (HOMA) method based on the formula HOMA{IR~Glucose(mg=dL)|Insulin(mU=mL)7405 [22].

Oxygen consumption and CO 2 production measurements
Oxygen consumption and CO 2 production were measured simultaneously for each mouse using a computer-controlled, Comprehensive Lab Animal Monitoring (CLAMS) System (Columbus Instruments, Columbus, OH) [23]. Each mouse was measured individually in a resting state for 24 hours at 22uC in presence of food and water or measured individually when running on a treadmill at a speed of 15 m/min for 40 min.

Measurement of blood inflammatory biomarkers
At the end of the study, blood was collected and plasma was stored at 280uC for the analysis of cytokines. Plasma levels of TNF-a, IFN-c, and monocyte chemoattractant protein 1 (MCP-1) were measured using Mouse Inflammation 6-Plex Kit from BD Bioscience (San Diego, CA), according to manufacturer's instructions. The cytokine levels were then determined using a BD LSR II instrument and analyzed by the BD CBA software (BD Biosciences, San Jose, CA) [21].

Transmission electron microscopy
To investigate the mitochondrial changes in situ between groups, we examined the ultrastructure of mitochondria by transmission electron microscopy (TEM) as previously described [19]. Briefly, muscle tissue was excised into small pieces (around 1 mm 3 ) and fixed in 2.5% gluteraldehyde (0.1 M phosphate buffer, pH 7.4) for 3 hours. Then each specimen was post-fixed in 1% osmium tetroxide for 1 hour and dehydrated through a graded ethanol series (50-100%). After embedded in eponate 12 resin, sections at a thickness of 80 nm were cut and stained by 2% aqueous uranyl acetate followed by lead citrate. The grids were loaded and observed in a Tecnai G2 Spirit transmission electron microscope (FEI, Hillsboro, OR). The images of mitochondria were captured at a magnification of 18,5006. For the morphometric analysis, five micrographs per tissue were counted. Mitochondrial size and number were analyzed by National Institutes of Health Image J software.

Data analyses
All data are expressed as means 6 SEM unless otherwise specified. Difference between two groups was tested by student's t test. Differences among groups were tested by two-way ANOVA and Boneferroni's post hoc test using GraphPad Prism ver. 5 (GraphPad Software, La Jolla, CA). P values of ,0.05 were considered statistically significant.

Results
Decreased expression of PKCb in both skeletal muscle and liver after exercise PKC proteins have been suggested to play a role in the insulin sensitivity of skeletal muscle [5,6,7]. To investigate the involve- ment of PKC in exercise, we tested the expression of different isoforms in the skeletal muscle and liver after 8 weeks of exercise. As depicted in Figures 1A & 1B, no significant differences of PKCa or d between exercise and sedentary groups were detected in both skeletal muscle and liver in WT mice. PKCl was increased in muscle and liver after exercise. PKCe level was slightly decreased in the muscle but was not changed in the liver of exercised mice compared with sedentary controls. It is noteworthy that PKCb was significantly reduced in both the skeletal muscle (SED 160.08 vs. EX 0.2160.09, p,0.05) and liver (SED 160.06 vs. EX 0.360.15, p,0.05) of exercised mice. These data suggested that PKCb might be involved in exercise-induced metabolic changes.  (Figure 2A). The total weight gain of sedentary WT mice was significantly higher than in the other groups, while no significant difference of weight gain was observed between exercised WT and exercised PKCb -/mice (WT SED 16.361.2 g vs. WT EX 11.961.7 g vs. PKCb -/-SED 12.461.7 g vs. PKCb -/-EX 8.261.6 g, p,0.05; Figure 2B). Despite lower body weight, the daily food and water intake of PKCb -/mice was similar to that of WT mice (Food intake: WT Exercise reduced fat mass and increased skeletal muscle weight in WT mice but not PKCb -/mice. Compared to WT controls, PKCb -/mice (both exercised and sedentary) had slightly increased tissue weights of skeletal muscles and significantly reduced tissue weights of inguinal fat, epididymal fat, and interscapular brown fat. The liver weight of sedentary WT mice was also higher than that of sedentary PKCb -/mice, while no significant difference of liver weight was observed among other groups ( Figure 3A). Similar results of tissue weight were obtained when normalized to body weight ( Figure 3B). Consistently, MRI scans also suggested that both the visceral and subcutaneous fat  Figures 4F & 4G). Similar plasma adiponectin levels were found among the four groups ( Figure 4H). Collectively, these results demonstrated PKCb deficiency enhances insulin sensitivity with no further improvement by exercise, suggesting exercise-induced improvements in insulin resistance may involve PKCb-mediated pathways.

Exercise increased metabolic rate via decrease in PKCb
As shown in Figures 5A-D Figures 5E-H). These results indicated that exercise increases metabolic rate possibly through decreasing PKCb.
Both exercise and PKCb deficiency reduced HFD-induced mitochondrial defects in the skeletal muscle As shown in Figure 6A, mitochondria from the muscle in exercised WT mice had a more clearly defined internal membrane Eight weeks of exercise did not have significant impact on adipose tissue inflammation Increased ratio of M1 (classically activated macrophages) versus M2 (alternatively activated macrophages) is suggested to be an important feature of adipose inflammation and insulin resistance [24]. To investigate the significance of exercise on adipose tissue inflammation, we detected the macrophage percentage and M1/ M2 ratio (CD11b + CD11c + cell/CD11b + CD204 + cell) in stromal vascular fraction (SVF) of epididymal fat. As shown in Figures 7A  & 7B, exercise slightly decreased macrophage numbers in epididymal SVF from WT mice although not to a statistically significant level. Deficiency of PKCb reduced macrophage infiltration in epididymal fat (WT SED 40.1364.29% vs. WT EX 36.4563.75% vs. PKCb -/-SED 26.3661.85% vs. PKCb -/-EX 22.0862.91%, p,0.05). Mice from all the four groups had similar levels of classical and alternative macrophage activation ( Figures 7C & 7D). In addition, the plasma levels of cytokines, including IL-6, IL-10, and MCP-1, were comparable among all the four groups ( Figures 7E-G).

Both exercise and PKCb deficiency enhanced insulin signaling in peripheral tissues
As depicted in Figure 8A, higher activation of AKT was observed in the liver of exercised WT mice, exercised PKCb -/mice, and sedentary PKCb -/mice when compared with that of sedentary WT mice. Similar results were found in the skeletal muscle ( Figure 8B).

Discussion
Obesity and diabetes are increasing worldwide at an alarming rate largely due to increased prosperity and sedentary life styles. Appropriate diets and exercise are two important interventions for both type 1 and type 2 diabetes. Characterizing the beneficial effects of exercise on insulin sensitivity has been the focus of the research. Despite important advancements in recent years, the mechanistic basis behind how exercise improves insulin signaling is still poorly understood. In the current study, we discovered a potential mechanism by which exercise improved insulin sensitivity in obese subjects.
Expression levels of several isoforms of PKC were altered in skeletal muscle of obese or diabetic patients [5,6,7]. Furthermore, forced expression of PKCb in skeletal muscle caused a decrease in activation of IRS1 and glucose uptake [25]. However, the involvement of PKC proteins in exercise and exercise-mediated metabolic changes is unknown. By detecting different isoforms of PKC, we found that PKCb levels were significantly decreased in both skeletal muscle and liver, suggesting that PKCb might be involved in exercise-mediated improvements in insulin sensitivity. Our study further demonstrated that PKCb deficiency, like exercise, could increase insulin sensitivity, as evidenced by improved ITT and HOMA-IR indices. Similarly, the activation of insulin downstream molecule AKT was enhanced by exercise and PKCb deficiency. Moreover, no significant differences in insulin sensitivity were observed between exercised and sedentary PKCb -/mice, indicating that exercise possibly attenuates insulin resistance via the reduction of PKCb levels. However, the response of exercised WT mice to glucose challenge in the IPGTT assay was not as significant as that in the ITT, although the blood glucose levels at 0 and 120 min were lower than those of sedentary WT mice. This result is not surprising given the dependence of the IPGTT response on a multitude of factors and this is likely caused by the compensation of insulin secretion, as plasma insulin levels of sedentary WT mice are significantly higher than those of exercised mice.
Consistent with the results reported by other groups [26,27], exercise lowered HFD-induced weight gain in WT mice. Exercised WT mice and PKCb -/mice had a lower fat mass than sedentary WT mice. Of note, no significant effects of exercise on weight gain and visceral fat mass were observed in PKCb -/mice. Compared to WT mice, PKCb -/mice had lower weight gain in response to HFD feeding, despite comparable food and water intake. This result suggested the PKCb may affect energy usage. Metabolic measurements indicated that both exercise and PKCb deficiency enhanced metabolic rate, whereas no effects of exercise on metabolic rate were found in PKCb -/mice. Despite less adiposity in PKCb -/mice, we did not observe a significant difference in the level of plasma adiponectin between WT and PKCb -/mice. This is probably caused by the fact that PKCb might have a suppressive effect on adiponectin expression. PKCb has been shown to induce the activation of JNK and subsequently suppress PPARc, a transcription factor promoting the expression of adiponectin [28,29].
In addition to insulin sensitivity and metabolism, exercise has beneficial effects on HFD-induced mitochondrial dysfunction, which has been observed in skeletal muscle of both obese rodent and human subjects [30,31,32]. Both an HFD and a high-sucrose diet can induce reactive oxygen species production in skeletal muscle, which results in mitochondrial dysfunction [30]. In this study, exercise reduced HFD-induced mitochondrial defects in WT mice. We also observed an ameliorated mitochondrial abnormality induced by HFD (including a more clearly defined internal membrane structure, and appropriate size and number) in both exercised and sedentary PKCb -/mice. It has been reported that adipose tissue inflammation contributes to the development of insulin resistance [33,34]. Recent studies suggest that long-term exercise reduces adipose inflammation via suppression of macrophage infiltration and a switch from M1 to M2 [35,36]. We detected a slight decrease in macrophage infiltration in exercised WT mice, although no statistical significance was found. We also failed to detect a switch from M1 to M2 in exercised mice, which was probably caused by a shorter exercise period and lower fat content (42%) than studied in previous reports [35,36]. However, PKCb deficiency significantly decreased macrophage infiltration in adipose tissue. Considering the fact that an exercise-induced decrease in PKCb levels in skeletal muscle does not have a significant effect on adipose tissue macrophage infiltration, the decrease in adipose tissue macrophages in PKCb -/mice likely resulted from a defect of PKCb in either adipose tissue or macrophages. In addition, exercise and PKCb deficiency did not affect the plasma levels of cytokines, including IL-6, IL-10, and MCP-1. These results suggested that 8 weeks of exercise may not have a significant effect on adipose inflammation and that exercise-induced improvements in insulin resistance may be unrelated to changes in adipose tissue macrophage content or function. Circulating level of IL-6 has been reported to be elevated after acute exercise. However, the effect of long-term exercise on IL-6 is controversial [37,38,39,40]. It's reported that no effect of long-term exercise on circulating IL-6 in elder adults, which is consistent with our observations [40,41].
Taken together, our results suggested that exercise decreased the expression of PKCb in both skeletal muscle and liver. By reducing PKCb expression, exercise improved HFD-induced metabolic dysfunction, including insulin resistance, fat accumulation, and mitochondrial dysfunction. Moreover, our findings that PKCb -/mice have an increased basal metabolic rate suggest PKCb could be a potential target for treating obesity and insulin resistance. However, the involvement of entire PKC family in exercise and insulin resistance might be complex due to the diversity of PKC isoforms and potential compensation among different isoforms. It requires further studies to investigate whether the other isoforms of PKC are involved in exercise-mediated improvement of insulin resistance. Figure S1 Exercise regimen. 4-week-old male WT and PKCb -/mice were fed a HFD for 16 weeks. At the age of 12 weeks, the mice were randomly assigned into 4 groups: WT exercise (WT EX); WT sedentary (WT SED); PKCb -/exercise (PKCb -/-EX); PKCb -/sedentary (PKCb -/-SED). Mice in EX group were exercise-trained on a motorized treadmill at a speed of 15 m/min, 40 min/day, 5 days/week for 8 weeks. Mice in SED group were put in the treadmill without running 40 min/day, 5 days/week for 8 weeks.