Exercise Training Induces Mitochondrial Biogenesis and Glucose Uptake in Subcutaneous Adipose Tissue Through eNOS-Dependent Mechanisms

  1. Roberto Vettor1
  1. 1Internal Medicine 3, Endocrine-Metabolic Laboratory, Department of Medicine DIMED, University of Padua, Padua, Italy
  2. 2Department of Biomedical Sciences, University of Padua, Padua, Italy
  3. 3Department of Molecular and Translational Medicine, University of Brescia, Brescia, Italy
  4. 4Center for Study and Research on Obesity, Department of Medical Biotechnology and Translational Medicine, University of Milan, Milan, Italy
  1. Corresponding author: Roberto Vettor, roberto.vettor{at}unipd.it.
  1. E.N. and R.V. contributed equally to this work.

Abstract

Insulin resistance and obesity are associated with a reduction of mitochondrial content in various tissues of mammals. Moreover, a reduced nitric oxide (NO) bioavailability impairs several cellular functions, including mitochondrial biogenesis and insulin-stimulated glucose uptake, two important mechanisms of body adaptation in response to physical exercise. Although these mechanisms have been thoroughly investigated in skeletal muscle and heart, few studies have focused on the effects of exercise on mitochondria and glucose metabolism in adipose tissue. In this study, we compared the in vivo effects of chronic exercise in subcutaneous adipose tissue of wild-type (WT) and endothelial NO synthase (eNOS) knockout (eNOS−/−) mice after a swim training period. We then investigated the in vitro effects of NO on mouse 3T3-L1 and human subcutaneous adipose tissue–derived adipocytes after a chronic treatment with an NO donor: diethylenetriamine-NO (DETA-NO). We observed that swim training increases mitochondrial biogenesis, mitochondrial DNA content, and glucose uptake in subcutaneous adipose tissue of WT but not eNOS−/− mice. Furthermore, we observed that DETA-NO promotes mitochondrial biogenesis and elongation, glucose uptake, and GLUT4 translocation in cultured murine and human adipocytes. These results point to the crucial role of the eNOS-derived NO in the metabolic adaptation of subcutaneous adipose tissue to exercise training.

Introduction

Reduced mitochondrial content and/or activity is associated with impaired cell function in several diseases (1,2). In particular, it has been hypothesized that mitochondrial impairment may be involved in the pathogenesis of obesity and insulin resistance and their progression toward type 2 diabetes (3,4), even though the role of mitochondria in health and disease is still under discussion (5,6). At the same time, metabolic disorders are also associated with a reduction of endothelial nitric oxide synthase (eNOS) enzymatic activity (7,8); in fact, mice lacking the eNOS gene are considered a useful murine model for metabolic syndrome because they display typical features, including hypertension, hypertriglyceridemia, endothelial dysfunction, insulin resistance, and visceral obesity (9).

It is well known that physical exercise induces profound physiological adaptations in several tissues as a response to increased metabolic requirements. One of the major events induced by physical activity is the upregulation of eNOS gene expression and the consequent increase of tissue nitric oxide (NO) production, which in turn induces mitochondrial biogenesis and cell glucose uptake in skeletal and cardiac muscle (1012). Moreover, a selective depletion of peroxisome proliferator–activated receptor-γ (PPAR-γ) coactivator 1α (PGC-1α), a key regulator of mitochondrial biogenesis, leads to a blunting of exercise-induced increases in mitochondrial respiratory chain proteins in muscle (13,14). Thus, a lack of response to physical exercise in mitochondrial biogenesis could be related to reduced or impaired NO metabolism.

Insulin sensitivity is also increased after a training period, and a single bout of exercise could enhance basal glucose uptake by increasing GLUT4 translocation to the cell membrane of skeletal (15) and cardiac (16) myocytes. Similar results have been observed in mouse and human tissues not directly involved in mechanical work and oxidative processes (17), but few studies have focused on subcutaneous adipose tissue. The capacity to oxidize fuel substrates to meet the energy demand is increased in tissues involved in contractile activity during exercise, such as heart and skeletal muscles; therefore, lipid and glucose uptake and oxidation need to be increased.

Emerging evidence suggests that these adaptations occur not only in working tissues, such as skeletal muscle and heart, but also in white adipose tissue (WAT) and brown adipose tissue, liver, brain, and kidney. Several studies show that exercise induces a “browning effect” in WAT by increasing mitochondrial protein activity (18) or brown adipocyte–specific gene expression (19). A recent study hypothesized that this effect might be mediated by irisin (FNDC5), a PGC-1α–dependent myokine secreted during physical activity that promotes thermogenesis and uncoupling processes in white adipocytes, whereas leptin and other key regulator genes of “white” development are downregulated (20). Despite numerous results supporting this hypothesis, the role of irisin is still controversial (21) and requires further investigations.

The aim of the current study was to investigate the mechanism underlying the response of subcutaneous adipose tissue to physical exercise in terms of mitochondrial biogenesis and glucose uptake. In particular, we explored whether NO can play a role in such metabolic adaptations.

Research Design and Methods

Mice and Exercise Protocol

Thirty-six adult (8 weeks old) male wild-type (WT; C57BL/6J) and eNOS−/− (B6.129P2-Nos3tm1Unc/J) mice (all from The Jackson Laboratory) were treated according to the European Union guidelines and with the approval of the institutional ethical committee. Body weight and food consumption were monitored throughout the experimental period. WT and eNOS−/− mice (n = 18 mice per group) were assigned randomly to swim training (12) or to have no lifestyle modifications. Mice swam once a day for 5 days/week in a graduated protocol starting at 10 min daily, with a 10-min increase each day until 90 min daily at the end of the second week. Then, mice swam 30 days on the full training regimen (90 min daily, 5 days/week). Swim sessions were supervised, water temperature was maintained between 30°C and 35°C, and mice were carefully towel-dried after each training session.

Muscle Contractile Performance In Vivo

Muscle strength developed by WT and eNOS−/− mice during instinctive grasp was measured with the protocol indicated as grip test (22). Briefly, the mouse was held by the tail near a trapeze bar connected to the shaft of a force transducer. Once the mouse had firmly grabbed the trapeze, a gentle pull was exerted on the tail. The measurement of the peak force generated by the mouse was repeated several times, with appropriate intervals to avoid fatigue, and average peak force values were expressed relative to body mass. Endurance was measured with a test to exhaustion on a treadmill. Initial speed (5 cm/s) was increased after 2 min at 10 cm/s. The speed was then increased by 2 cm/s every minute up to 50 cm/s, and time to exhaustion was recorded.

Tissue Glucose Utilization Index

At the end of the training period and 2 days before the clamp studies, a catheter was inserted into the right femoral vein under general anesthesia with sodium pentobarbital. Tissue glucose uptake studies were performed on mice under conscious and unstressed conditions after an 8-h fast. As previously described, with minor modifications (23), 10 μCi of the nonmetabolizable glucose analog 2-deoxy-d-[1-3H]glucose ([3H]-DG) (Amersham Biosciences) was injected as an intravenous bolus in the basal condition or after a hyperinsulinemic euglycemic clamp. Animals were killed 120 min after the tracer injection, and subcutaneous adipose tissue from the inguinal fat pad was quickly collected in liquid nitrogen and kept at −80°C for subsequent analysis. The glucose utilization index was derived from the amount of [3H]-DG-6-phosphate ([3H]-DGP) measured in adipose tissue as previously described (24), thus using the accumulation of [3H]-DGP as an index of the glucose metabolic rate.

Norepinephrine Treatment

A group of sedentary mice (n = 10 WT and n = 10 eNOS−/−) of the same strains as those used for the training experiments received a single intraperitoneal injection of norepinephrine hydrogen tartrate (5 mg/kg; Galenica Senese) or vehicle (0.9% saline solution). Animals were killed 24 h after the norepinephrine injection, and the subcutaneous inguinal fat pad was quickly removed and kept at −80°C for subsequent mRNA expression analysis. An additional group of sedentary mice (n = 10 WT and n = 10 eNOS−/−) received the same norepinephrine treatment and were killed 30 min after the intraperitoneal injection. The inguinal fat pad was quickly removed and kept at −80°C to quantify protein expression and phosphorylation.

Immunoblot Analysis

Proteins were isolated from inguinal subcutaneous adipose tissues using T-PER Mammalian Protein Extraction Reagent (Pierce), as indicated by the manufacturer, in the presence of 1 mmol/L NaVO4, 10 mmol/L NaF, and a cocktail of protease inhibitors (Sigma-Aldrich). Protein content was determined by the bicinchoninic acid protein assay (Pierce), and 70 µg proteins were run on SDS-PAGE under reducing conditions. The separated proteins were then electrophoretically transferred to a nitrocellulose membrane (Pierce). Proteins of interest were revealed with specific antibodies: anti-GLUT4, anti-AKT, anti–phospho-AKT, anti-p44/42 mitogen-activated protein kinase (MAPK) extracellular signal-regulated kinases 1 and 2 (ERK1/2), anti–phospho-ERK1/2, anti–hormone-sensitive lipase (HSL), anti–phospho-HSL (all from Cell Signaling), and anti–β-actin (Sigma-Aldrich). The immunostaining was detected using horseradish peroxidase-conjugated anti-rabbit or anti-mouse Ig for 1 h at room temperature. Bands were revealed by the SuperSignal Substrate (Pierce) and quantified by densitometry using ImageJ software (National Institutes of Health).

Cell Cultures and Treatment

Stromal vascular fraction from subcutaneous adipose tissue of five healthy patients undergoing bariatric surgery was isolated as previously described (25). Human-derived and 3T3-L1 (ATCC CL-173) preadipocytes were plated in Dulbecco's modified Eagle's medium, supplemented with 10% FBS, 150 units/mL streptomycin, 200 units/mL penicillin, 2 mmol/L glutamine, and 1 mmol/L HEPES (all from Life Technologies). At confluence, adipogenic differentiation was induced by adding 1 μmol/L dexamethasone, 0.5 mmol/L 3-isobutyl-1-methyl-xantine (IBMX, Sigma-Aldrich), and 70 nmol/L insulin (Novo Nordisk). IBMX was removed from medium after 3 days of culture. Cells were cultured at 37°C in a humidified atmosphere of 5% CO2 until fully differentiated. Cells were exposed to 100 μmol/L diethylenetriamine-NO (DETA-NO; Sigma-Aldrich), a potent NO donor (26), or vehicle for 72 h and then harvested for further analysis.

Cell Glucose Uptake

Human-derived and 3T3-L1 mature adipocytes were serum-starved for 8 h, incubated in the presence or absence of 2 μmol/L insulin (Novo Nordisk) for 30 min, and then with 1.5 μCi/mL [3H]-DG (Amersham Biosciences) for 15 min. Cells were washed with ice-cold PBS and lysed in 0.5 mol/L NaOH. Radioactivity was measured by liquid scintillation counting (Wallac). Each condition was assayed in three independent experiments in triplicate.

RNA Isolation and Real-Time Quantitative PCR

Total RNA was isolated from cultured adipocytes and frozen subcutaneous adipose tissues of mice using RNeasy Mini kit or RNeasy Lipid Tissue Mini Kit (Qiagen), respectively, treated with DNase (TURBO-DNase-free, Ambion), and reverse-transcribed using random primers (Promega) and M-MLV reverse transcriptase (Promega). mRNA levels were measured by real-time quantitative PCR (qPCR) (DNA Engine Opticon2, MJ Research) using SYBR Green PCR Master Mix (Applied Biosystems) and specific intron-spanning primers according to the manufacturer’s instructions. All data were collected in triplicate and normalized to 18S gene expression.

Mitochondrial DNA

Mitochondrial DNA (mtDNA) copy number was measured by means of qPCR from the cytochrome B mtDNA gene compared with the large ribosomal protein p0 (36B4) nuclear gene, as previously described (27).

Mitochondrial Morphology and Membrane Potential

After DETA-NO treatment, cells were incubated with 100 nmol/L of MitoTracker Green FM (MTG) dye (Molecular Probes) or with 5 μmol/L JC-1 (Molecular Probes) for 30 min at 37°C and 5% of CO2. After a 37°C medium wash, cells were observed using a Nikon Ti-E equipped with DS-2M cooled camera (Nikon) and a top-stage incubator Tokai Hit INU (Tokai Hit) to maintain optimal culturing conditions. MTG staining was evaluated by fluorescence imaging (490/516 nm) with a ×60 1.45 numerical aperture objective, and JC-1 staining was evaluated (514/529 nm for the monomeric state to 585/590 nm as dimeric state) with a ×100 0.75 numerical aperture objective.

Digital Imaging Processing

After acquisition, images were processed with NIS-Elements AR software (Nikon) to evaluate the shape and area of MTG-stained cells. Briefly, binary objects were obtained by image segmentation; then, an automatic measurement tool was used to calculate the area and elongation (intended as the maximum Feret-to-minimum Feret diameter ratio, where Feret diameter is the distance between the two parallel planes restricting the object perpendicular to that direction) of each mitochondria.

GLUT4 Staining

After 2 μmol/L insulin stimulation for 30 min, cells were fixed on coverslips with 4% paraformaldehyde and permeabilized with 0.25% Triton X-100. GLUT4 was detected by incubation with a polyclonal primary antibody (Santa Cruz Biotechnology) using standard procedures. After washing with PBS, binding of primary antibodies was detected with Alexa-488–conjugated secondary antibodies (Life Technologies). Coverslips were mounted with ProLong Gold antifade medium (Life Technologies) and analyzed with the confocal Leica DMI6000 CS SP8 laser scanning microscope (Leica Microsystem). Image analysis was performed with ImageJ software.

Statistical Analysis

Results are expressed as means ± SEM. Data were analyzed by GraphPad Prism 5.0 software using unpaired Student t test, one-way ANOVA, or two-way ANOVA with Newman-Keuls post hoc test, as appropriate.

Results

eNOS Is Required for Mitochondrial Biogenesis in Response to Exercise in Subcutaneous Adipose Tissue

Swim training lasting 6 weeks was selected to induce mitochondrial biogenesis in WT mice and in eNOS−/− mice. No significant differences in muscle strength, as measured with grip tests, were present between sedentary WT and eNOS−/− mice, although slightly higher values were observed in trained animals (Table 1). Importantly, the training was able to significantly improve endurance in WT but not in eNOS−/− mice. To determine if swim training induces an increase in mitochondrial biogenesis and/or function in adipose tissue and whether such an increase is NO-dependent, we measured the mRNA level of multiple members of mitochondrial transcriptional machinery as well as mtDNA content (an index of mitochondrial mass) in the subcutaneous adipose tissue of WT and eNOS−/− mice. The expression of PGC-1α, nuclear respiratory factor 1 (NRF-1), mitochondrial transcription factor A (Tfam), and cytochrome c oxidase IV (COX IV) mRNA was increased in subcutaneous adipose tissue of WT trained mice compared with sedentary control littermates (Fig. 1A).

Table 1

Effect of swim training on mice performance evaluated in vivo

Figure 1

Mitochondrial biogenesis in mouse subcutaneous adipose tissue. WT and eNOS−/− mice (n = 8 per group) were swim-trained for 6 weeks, killed, and subcutaneous adipose tissue was collected from the inguinal fat pad. A: Relative mRNA levels were measured by combined reverse transcription (RT) and qPCR techniques (qRT-PCR) using 18s rRNA as the internal control and expressed as fold change. B: mtDNA content was measured by means of real-time PCR and expressed as % of mtDNA copy number per nuclear DNA copy number. C: Representative Western blots show PGC-1α, COX IV, and β-actin immunodetected signals in WAT lysates of mice. SED, sedentary; D: Protein expression levels were measured by Western blot analysis using β-actin as the internal control and are expressed as fold change. All graphs depict mean ± SEM. Two-way ANOVA, *P < 0.05 and **P < 0.01 relative to sedentary mice; †P < 0.05 and ††P < 0.01 relative to WT mice.

In contrast, eNOS−/− mice, which showed reduced mRNA content of PGC-1α and COX IV in untrained conditions compared with WT mice, failed to display the mitochondrial biogenic response to exercise after the swim training period (Fig. 1A). The “browning” effect of exercise training on WAT was confirmed by the increase of UCP-1 mRNA content in trained WT animals, whereas the basal content of UCP-1 in eNOS−/− mice was very low and did not increase after exercise training. Interestingly, we also measured a basal mRNA expression of FNDC5, which is the precursor of the newly discovered hormone irisin (20), in subcutaneous adipose tissue of WT mice. FNDC5 was higher in WT compared with eNOS−/− sedentary mice and did not change after exercise training. Furthermore, swim training increased mtDNA content of subcutaneous adipose tissue of WT but not eNOS−/− mice (Fig. 1B).

We observed a significant increase in PGC-1α and COX IV protein expression in subcutaneous adipose tissue of WT but not eNOS−/− mice after the swim training protocol. Moreover, eNOS−/− mice displayed a significantly lower PGC-1α protein content in trained conditions compared with WT trained animals (Fig. 1C and D).

To ascertain if modifications in β-adrenergic sensitivity could have influenced our findings, the effect of one intraperitoneal injection of norepinephrine (5 mg/kg) was assessed in subcutaneous adipose tissue of WT and eNOS−/− mice. We observed a significant increase in PGC-1α and COX IV mRNA levels in WT mice 24 h after the norepinephrine treatment (Fig. 2A). This effect was present also in eNOS−/− mice but to a remarkably lesser extent. Moreover, a slight increase in Tfam and NRF-1 mRNA levels was observed after β-adrenergic stimulation in WT but not in eNOS−/− mice (Fig. 2A). To further demonstrate that the attenuated norepinephrine-induced expression of PGC-1α and other mitochondrial biogenesis factors in adipose tissue of eNOS−/− mice is due to a lack of eNOS and not to an impaired β-adrenergic signaling, we evaluated the phosphorylation state of two protein kinase A (PKA) substrates, ERK1/2 and HSL, 30 min after the adrenergic stimulation. We observed that phospho-ERK1, phospho-ERK2, and phospho-HSL protein levels were significantly higher after norepinephrine injection in WAT of WT and eNOS−/− mice compared with untreated control animals (Fig. 2B and C). Notably, we did not observe any significant difference in basal protein levels of ERK1/2 or HSL between WT and eNOS−/− mice (Fig. 2B and C). These results suggest the functional integrity of the β-adrenergic signaling pathway in eNOS-null mutant mice.

Figure 2

Norepinephrine-induced mitochondrial biogenesis and β-adrenergic signaling in mouse subcutaneous adipose tissue. WT and eNOS−/− mice (n = 10 per group) received one intraperitoneal injection of 0.9% saline solution (control [CTRL]) or norepinephrine (5 mg/kg), were killed, and subcutaneous adipose tissue was collected from inguinal fat pad. A: Relative mRNA levels were measured by qRT-PCR in WAT lysates of mice killed 24 h after injection, using 18s rRNA as the internal control and expressed as fold change. B: Representative Western blots show HSL, phospho (P)-HSL, ERK1/2, phospho (P)-ERK1/2, and β-actin immunodetected signals in WAT lysates of mice killed 30 min after the injection with saline (CTRL) or norepinephrine (TREATED) solution. C: Protein expression was measured by Western blot analysis using β-actin as the internal control. The signals obtained from phosphorylated proteins were normalized, each one to the corresponding total protein level. All graphs depict mean ± SEM. Two-way ANOVA, *P < 0.05 relative to control mice; †P < 0.05 relative to WT mice.

To rule out a possible nonspecific effect caused by the stress elicited by the intraperitoneal injection, we compared the previous results with those obtained in a control group of WT and eNOS−/− mice that did not receive any injection. We did not observe any statistically significant difference between treated and untreated mice (data not shown).

eNOS Is Required for Basal and Insulin-Stimulated Glucose Uptake in Subcutaneous Adipose Tissue in Response to Exercise

To explore if swim training improves glucose metabolism or insulin sensitivity in subcutaneous adipose tissue of eNOS−/− mice, we measured the tissue glucose utilization index as the amount of [3H]-DG uptake in the basal condition and after a hyperinsulinemic euglycemic clamp. Swim training increased basal and insulin-stimulated glucose uptake of subcutaneous adipose tissue in WT but not eNOS−/− mice (Fig. 3A). Moreover, eNOS−/− mice displayed a lower glucose uptake capacity, even in the sedentary condition, compared with WT control mice (Fig. 3A). We observed a significant increase in pAKT-to-AKT protein content in subcutaneous adipose tissue of WT trained mice compared with sedentary littermates (Fig. 3B and C). The pAKT-to-AKT ratio did not increase in the subcutaneous adipose tissue in eNOS−/− after the swim protocol, and they showed a significantly lower protein expression in trained conditions than trained WT animals. We also measured the GLUT4 total protein content in subcutaneous adipose tissue of sedentary and trained animals, but no significant changes were observed in WT or eNOS−/− mice (Fig. 3B and C).

Figure 3

Glucose uptake and insulin signaling in mouse subcutaneous adipose tissue. A: For glucose uptake measurement, mice (n = 8 per group) were fasted 8 h and injected with PBS (left) or insulin (0.5 units/kg body weight, right). B: Representative Western blots show AKT, phospho (P)-AKT, GLUT4, and β-actin immunodetected signals in protein lysates obtained from WAT of sedentary (SED) and trained WT and eNOS−/− mice. C: Protein expression levels were measured by Western blot analysis using β-actin as the internal control and expressed as fold change. All graphs depict mean ± SEM. Two-way ANOVA, *P < 0.05 relative to sedentary mice; †P < 0.05 relative to WT mice.

NO Promotes Mitochondrial Biogenesis and Elongation, Glucose Uptake, and GLUT4 Translocation in 3T3-L1 Adipocytes

To determine if an increased NO bioavailability could improve mitochondrial biogenesis and/or glucose uptake in mouse adipocytes, we first measured mRNA content of PGC-1α, NRF1, Tfam, and COX IV in fully differentiated 3T3-L1 cells after a 72-h treatment with 100 μmol/L DETA-NO. Gene expression of mitochondrial biogenesis markers was upregulated in DETA-NO–treated cells compared with untreated control cells (Fig. 4A). We then evaluated the mitochondrial morphometric parameters by MTG labeling, followed by a digital image analysis able to measure the area and the elongation of each mitochondrion (see digital imaging processing for details). We observed a higher presence of larger (area >1.5 μm2) and longer (length >3 μm) mitochondria in treated cells compared with those observed in untreated control cells (Fig. 4BD). Glucose uptake was measured after DETA-NO treatment by performing an in vitro [3H]-DG uptake assay. We observed that DETA-NO increased glucose transport of 3T3-L1 cells in basal and in insulin-stimulated (2 μmol/L) conditions (Fig. 5A). Moreover DETA-NO treatment induced an increase in GLUT4 translocation to the cell membrane (measured as cytoplasm-to-membrane fluorescence ratio) in basal and insulin-stimulated conditions compared with untreated control cells (Fig. 5B).

Figure 4

Mitochondrial biogenesis in 3T3-L1 adipocytes. Fully differentiated 3T3-L1 adipocytes were treated for 72 h with 100 μmol/L vehicle (CTRL) or 100 μmol/L DETA-NO. A: mRNA levels were analyzed by means of qRT-PCR using 18s rRNA as the internal control and expressed as fold change (n = 3 independent experiments). B: MTG dye was used as an indicator of mitochondrial mass in live cells. Mitochondrial area (C) and elongation (D) were analyzed as the percentages of larger (>1.5 μm2) and longer (>3 μm) mitochondria in DETA-NO–treated cells compared with untreated cells (n = 3 independent experiments). Data are expressed as mean ± SEM. Student t tests, *P < 0.05 and **P < 0.01 relative to untreated cells (CTRL).

Figure 5

Glucose uptake and GLUT4 translocation in 3T3-L1 adipocytes. Fully differentiated 3T3-L1 adipocytes were treated for 72 h with vehicle (CTRL) or 100 μmol/L DETA-NO. Cells were serum-starved for 8 h, treated with PBS (basal) or 2 μmol/L insulin for 30 min and subjected to a [3H]-DG uptake assay (A) or immunostained with anti-GLUT4 antibody (B). Fluorescence was detected by confocal microscopy, and ratio between the membrane and cytoplasmic signal was measured by means of image analysis. All data are expressed as mean ± SEM (n = 3 independent experiments). Two-way ANOVA, *P < 0.05 and **P < 0.01 relative to control cells; †P < 0.05 relative to cells in basal conditions.

NO Promotes Mitochondrial Biogenesis and Polarization in Human Subcutaneous Adipocytes

To further investigate the role of NO in adipose tissue, DETA-NO effects were also studied in subcutaneous adipocytes obtained from abdominal biopsy specimens of subjects undergoing bariatric surgery. The stromal vascular fraction of adipose tissue was isolated and adipocytes cultured to full differentiation in vitro. We observed that PGC-1α, Tfam, COX IV, and ATP-synthase subunit-β (ATPS) gene expression was upregulated in human mature adipocytes treated with DETA-NO compared with untreated cells (Fig. 6A). Moreover, DETA-NO treatment increased mtDNA content in human mature adipocytes (Fig. 6B). We then labeled the adipocyte mitochondria with the vital staining dye JC-1 and observed that mitochondrial membrane potential was higher in cells after DETA-NO treatment compared with mitochondrial membrane potential of untreated cells (Fig. 6C).

Figure 6

Mitochondrial biogenesis and membrane potential in human adipocytes. Fully differentiated human adipocytes isolated from subcutaneous adipose tissue were treated for 72 h with control (CTRL) vehicle or 100 μmol/L DETA-NO. A: mRNA levels were analyzed by qRT-PCR using 18s rRNA as the internal control and are expressed as fold change (n = 3 independent experiments). B: mtDNA content was measured by means of qPCR and expressed as % of mtDNA copy number per nuclear DNA copy number (n = 3 independent experiments). Data are expressed as mean ± SEM. Student t tests, *P < 0.05 and **P < 0.01 relative to control cells. C: JC-1 assay was used as an indicator of mitochondrial membrane polarization in live cells. A representative image from three independent experiments is shown.

NO Promotes Glucose Uptake and GLUT4 Translocation in Human Subcutaneous Adipocytes

Glucose uptake and GLUT4 translocation were measured in fully differentiated human adipocytes as described for 3T3-L1 cells. We observed an increase in basal and insulin-stimulated glucose uptake capacity in DETA-NO–treated cells compared with untreated cells (Fig. 7A). Moreover, the fraction of GLUT4 translocated on cells membrane was significantly higher after DETA-NO treatment, even in the basal condition, and slightly increased in insulin-stimulated conditions (Fig. 7B and C).

Figure 7

Glucose uptake and GLUT4 translocation in human subcutaneous adipocytes. Fully differentiated human adipocytes isolated from subcutaneous adipose tissue were treated for 72 h with control (CTRL) vehicle or 100 μmol/L DETA-NO. Cells were serum-starved for 8 h, treated with PBS (basal) or 2 μmol/L insulin for 30 min, and subjected to a [3H]-DG uptake assay (A) or immunostained with anti-GLUT4 antibody (B). C: Fluorescence was detected by confocal microscopy, and the ratio between the membrane and cytoplasmic signal was measured by means of image analysis. All data are expressed as mean ± SEM (n = 3 independent experiments). Two-way ANOVA, *P < 0.05 and **P < 0.01 relative to control cells; †P < 0.05 relative to cells in basal conditions.

Discussion

The current study provides strong evidence that NO generated by eNOS plays a crucial role in the mitochondrial biogenesis and metabolic activation taking place in adipocytes in response to physical activity. A reduction of mitochondrial abundance and activity has been linked with insulin resistance in obesity and type 2 diabetes (28). One of the most important physiological conditions widely recognized to increase insulin sensitivity and mitochondrial biogenesis is physical activity: the common thought that exercise increases mitochondrial biogenesis mostly, if not exclusively, in skeletal muscles has recently been replaced by the notion that exercise training can induce mitochondrial biogenesis in a wide range of tissues not normally associated with the metabolic demand of exercise (29). Much emphasis has been placed on the role of mitochondrial biogenesis in the adipose organ in relation to its potential shift from the white to brown phenotype or to the activation of the beige phenotype. This phenomenon has been shown to be associated with increased energy expenditure and whole-body insulin sensitivity (30,31). In this study, we report that physical training in mice was able to induce an increase in subcutaneous adipose tissue mRNA and protein levels of key transcriptional regulators of mitochondrial biogenesis, along with an increase in mtDNA content. Interestingly, at the end of the training period, an increased WAT mRNA expression of UCP1, the specific marker of brown adipose tissue, was also observed in subcutaneous fat.

These results confirm and extend what was previously observed by Sutherland et al. (32), who reported that exercise training increased PGC-1α and Tfam mRNA expression, as well as COX IV protein and citrate synthase activity in rat visceral fat depots, further supporting the evidence that exercise training is an effective stimulus for mitochondrial biogenesis in WAT. This phenomenon has been explained by the sympathetic stimulation during exercise, which has been shown to induce PGC-1α in several tissues, including brown and WAT. Adrenaline directly stimulates PGC-1α in WAT, and the β-blocker propranolol is able to reduce the exercise-induced rise in PGC-1α by ∼40% (32). Among the mechanisms responsible for the remaining ∼60% of the exercise effect on WAT mitochondria biogenesis, an increased production of myokines, hormones (i.e., thyroid hormones, glucocorticoids), signaling molecules coming from the exercising muscles (i.e., irisin, NO), or other still unknown factors could be taken into consideration. The results collected in this study add a novel and important piece of evidence showing that a significant portion of the induction in mitochondrial biogenesis in WAT after training is due to activation of eNOS system. We observed a reduced basal expression of genes regulating mitochondrial biogenesis in adipose tissue of eNOS−/− mice and, more importantly, the physical training period failed to induce any increase in these transcriptional regulators or in mtDNA content. The sympathetic stimulation of brown adipose tissue in vivo and in vitro significantly increased eNOS expression and activity (33), and we observed that the norepinephrine-induced upregulation of mitochondrial biogenesis factors was reduced in WAT of eNOS−/− mice.

To evaluate any changes of the sympathetic activity in the knockout mice, we investigated the noradrenergic signaling in these animals. In adipose tissue, HSL enzyme activity is strongly induced by β-adrenergic stimulation. HSL and ERK are major targets for PKA-mediated phosphorylation, the first step downstream of the β-adrenergic receptor. ERK also phosphorylates HSL to modulate the activity of the enzyme. We observed that the norepinephrine-induced phosphorylation of ERK1/2 and HSL were not impaired in WAT of eNOS−/− mice. These results suggest that the eNOS system may play a role in mediating the adrenergic effect on mitochondrial biogenesis activation also in WAT.

Apparently, an increase in UCP1 expression in WAT may be in contrast with the known effects of exercise on mitochondrial coupling efficiency in skeletal muscle (that increases to produce more ATP), but this phenomenon could be explained by considering the different metabolic functions of skeletal muscle and adipose tissue, which imply different regulatory mechanisms of mitochondrial respiration and energy expenditure (34). Moreover, it has been observed that physical exercise could induce an increase of uncoupling activity also in skeletal muscle (35,36); therefore, the balance between substrates oxidation, membrane potential, and ATP production is a complex phenomenon that may have controversial aspects.

The role of NO in the activation of mitochondrial biogenesis and the regulation of skeletal muscle glucose uptake during exercise have been extensively studied in humans and in rodents (37), but so far, little data have been available regarding the effect of exercise on adipose tissue insulin sensitivity and the role of NO in mediating this effect (38,39). The data presented here clearly show that the significant increase in the exercise-induced adipose tissue glucose uptake in basal conditions and after insulin infusion in WT mice is abolished in eNOS−/− mice. Therefore, NO mediates, at least partially, the insulin-sensitizing effect of exercise on subcutaneous adipose tissue. To further clarify whether the observed effect of exercise on adipose tissue mitochondrial biogenesis and insulin sensitivity could be a genuine consequence of eNOS activation, irrespective of the stimulation of fat cell β-adrenoceptors (40,41), we studied the in vitro effect of an NO donor on murine and human fully differentiated adipocytes. The obtained results allowed us to confirm the role of NO in inducing the master genes regulating mitochondrial biogenesis and function, along with a parallel increase of basal and insulin-stimulated GLUT4 translocation and glucose uptake in mouse and human white adipocytes. Our findings highlight the important role of mitochondria in the regulation of adipocyte glucose homeostasis and support recent observations obtained after induction of mitochondrial dysfunction by knocking down Tfam, which led to a downregulation of GLUT4 expression and to an attenuation of insulin-stimulated glucose uptake in 3T3-L1 adipocytes (42). Accordingly, the treatment with PPAR-γ agonists improves the dysfunctional adipose organ by increasing white adipocyte insulin sensitivity and mitochondrial biogenesis in WAT of insulin-resistant mice (43).

Morphological and structural changes of mitochondria—from small fragmented units to larger networks of elongated organelles—play a role in several cellular processes according to the functional status of the cell. It has been recently reported that mitochondrial elongation is critical to sustain cell viability during macroautophagy induced by nutrient restriction (44). Longer mitochondria are protected from being degraded and possess more cristae where activity of the ATP synthase is increased, optimizing ATP production in times of nutrient restriction (44). Mitochondria unable to elongate during nutrient deprivation consume cellular ATP, leading to cell death. On the contrary, if elongation is blocked, mitochondria become dysfunctional and “cannibalize” cytoplasmic ATP to maintain their membrane potential, precipitating cell death (45). Whenever energy is needed, as in prolonged exercise (which mimics what happens during nutrient restriction), morphological changes of mitochondrial shape occur (46,47).

We aimed to determine if an increased NO availability could influence mitochondrial remodeling in adipocytes and found that chronic treatment with DETA-NO was able to induce a significant increase in mitochondrial area and promoted mitochondrial elongation in 3T3-L1 cells, possibly contributing to improved bioenergetics. We also observed that the transcriptional machinery of mitochondrial biogenesis was activated in DETA-NO–treated cells, along with an increased COX IV and ATP-synthase mRNA expression and increased mitochondrial potential. In contrast with the in vivo studies, we observed only a slight, not significant increase in UCP1 mRNA expression in white adipocytes treated with the NO donor. This could be explained considering the in vivo situation, where we observed the results of a complexity of synergistic phenomena occurring during exercise, including the prolonged sympathetic overdrive. On the contrary, the in vitro environment probably excludes some major determinants that play a role in mediating the exercise effect on adipose tissue in vivo, even if it represents a useful model to assess the importance of NO per se in influencing mitochondrial biogenesis and glucose handling by fat cells. The role of exercise in preventing obesity-related metabolic disorders by acting on adipose tissue has been recently highlighted (48,49). The effect of exercise is not limited to the exercising muscles, increasing fatty acids mobilization to match the peripheral energy requirement, but deeply influences the whole adipocyte energy metabolism by activating mitochondrial biogenesis and promoting insulin sensitivity (50). Our data show that these two phenomena are tightly associated and point to a crucial role of the eNOS system in mediating the effect of exercise training on the bioenergetic adaptation in subcutaneous white fat.

Article Information

Funding. This study was supported by Ministero dell'Università e della Ricerca (grants 20075HJTHM_003 to A.V., 2007HJTHM_001 to E.N., and 20112010329EKE_005 to R.V.), Ministero della Salute (grant RF-2009-1526404 to R.V.), and the Cariparo Foundation.

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

Author Contributions. E.T., M.S., M.O., M.G., L.T., R.F., R.S., and M.Q. performed research. C.R. and E.N. contributed new reagents and analytic tools. E.T., M.S., and A.V. analyzed data. E.T., A.V., C.R., E.N., and R.V. wrote the manuscript. E.N. and R.V. designed the research. E.N. and R.V. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Footnotes

  • M.Q. is currently affiliated with the Department of Neurology and Neurological Sciences, School of Medicine, Stanford University, Stanford, CA.

  • See accompanying article, p. 2606.

  • Received August 12, 2013.
  • Accepted March 9, 2014.

References

| Table of Contents

This Article

  1. Diabetes vol. 63 no. 8 2800-2811
  1. All Versions of this Article:
    1. db13-1234v1
    2. db13-1234v2
    3. db13-1234v3
    4. 63/8/2800 most recent