Arsenic Induces Insulin Resistance in Mouse Adipocytes and Myotubes Via Oxidative Stress-Regulated Mitochondrial Sirt3-FOXO3a Signaling Pathway

Abstract

Chronic exposure to arsenic via drinking water is associated with an increased risk for development of type 2 diabetes mellitus (T2DM). This study investigates the role of mitochondrial oxidative stress protein Sirtuin 3 (Sirt3) and its targeting proteins in chronic arsenic-induced T2DM in mouse adipocytes and myotubes. The results show that chronic arsenic exposure significantly decreased insulin-stimulated glucose uptake (ISGU) in correlation with reduced expression of insulin-regulated glucose transporter type 4 (Glut4). Expression of Sirt3, a mitochondrial deacetylase, was dramatically decreased along with its associated transcription factor, forkhead box O3 (FOXO3a) upon arsenic exposure. A decrease in mitochondrial membrane potential (Δψm) was observed in both 3T3L1 adipocytes and C2C12 myotubes treated by arsenic. Reduced FOXO3a activity by arsenic exhibited a decreased binding affinity to the promoters of both manganese superoxide dismutase (MnSOD) and peroxisome proliferator-activated receptor-gamma coactivator (PGC)-1α, a broad and powerful regulator of reactive oxygen species (ROS) metabolism. Forced expression of Sirt3 or MnSOD in mouse myotubes elevated Δψm and restored ISGU inhibited by arsenic exposure. Our results suggest that Sirt3/FOXO3a/MnSOD signaling plays a significant role in the inhibition of ISGU induced by chronic arsenic exposure.

Human exposure to arsenic via drinking water is a major public health concern due to toxicity and carcinogenicity of this metal. Millions of people worldwide are exposed to arsenic by contaminated drinking water (Jones, 2007). Globally in 2010, 285 million people suffered from type 2 diabetes mellitus (T2DM). It is estimated that the incidence of T2DM will increase to 439 million by 2030 (Shaw et al., 2010). In this scenario, research addressing the role of environmental chemicals in T2DM is becoming an important environmental health issue and has been rapidly expanded in the previous few years (Becker and Axelrad, 2014; Douillet et al., 2013; Fu et al., 2010; Huang et al., 2014; Li et al., 2013; Maull et al., 2012; Xue et al., 2011). Evidences from in vitro, in vivo, and population studies indicate that arsenic or its metabolites impair insulin-dependent glucose uptake, leading to insulin resistance (Paul et al., 2007; Walton et al., 2004; Xue et al., 2011). However, diabetogenic role of arsenic needs to be confirmed and the mechanisms of arsenic-induced diabetes are still under investigation.

Arsenic-induced oxidative stress is considered to be important in the mechanisms of diabetes induced by arsenic (Fu et al., 2010; Maull et al., 2012; Navas-Acien et al., 2008; Paul et al., 2007; Pi et al., 2003a,b; Tseng, 2004; Xue et al., 2011). Arsenic influences mitochondrial membrane potential (Δψm) and lysosomal membrane stability directly or through generation of reactive oxygen species (ROS) and related oxidative damage (Boya et al., 2003). The mechanism on how chronic low level arsenic exposure causes disruption of insulin stimulated glucose uptake (ISGU) and thus insulin resistance in insulin sensitive tissues such as adipocytes and myotubes remains to be studied. Previous studies have suggested that arsenic might impair pancreatic β-cell functions, mainly insulin synthesis and secretion (Díaz-Villaseñor et al., 2006; Woods et al., 2009). Subchronic exposure to arsenic induces oxidative stress and damages pancreas, which could be implicated as a cause of insulin resistance due to activation of stress-sensitive signaling pathways that involve transcription factors such as NF-κB (Flora, 2011; Izquierdo-Vega et al., 2006).

Sirtuin-3 (Sirt3), a mitochondria-targeted deacetylase, predominantly expresses in highly metabolic tissues and has been shown to bind and deacetylate several metabolic and respiratory enzymes that regulate important mitochondrial functions (Hirschey et al., 2010; Onyango et al., 2002; Someya et al., 2010; Tao et al., 2010). Dysfunction of mitochondria has been linked with diabetes and insulin resistance (Shi et al., 2005). Previous studies have demonstrated possible link between Sirt3 and T2DM (Boyle et al., 2013; Caton et al., 2013; Hirschey et al., 2010; Jing et al., 2011, 2013; Sundaresan et al., 2009). Mitochondrial Sirt3 induces forkhead box O3 (FOXO3a) translocation to the nucleus and augments FOXO3a-dependent antioxidant defense mechanism, through upregulation of manganese superoxide dismutase (MnSOD) (Sundaresan et al., 2009). Peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC-1α) stimulates mitochondrial biogenesis and electron transport activity. It suppresses ROS production and protects cells from oxidative stressor-induced death through induction of several key ROS-detoxifying enzymes (Kong et al., 2010). Insulin-regulated glucose transporter type 4 (Glut4), an insulin sensitive transporter characteristic of adipose tissue and skeletal muscle, is responsible for ISGU, a key process for the normalization of postprandial blood glucose levels.

This study investigates the mechanism of reduced ISGU by chronic arsenic exposure, focusing on activities of Sirt3 protein and its downstream targets such as FOXO3a, MnSOD, and PGC-1α in differentiated 3T3L1 adipocytes and C2C12 myotubes.

MATERIALS AND METHODS

Chemicals and reagents

Dulbecco’s modified Eagle’s medium (DMEM), fetal bovine serum (FBS), phosphate buffered saline (PBS), TRIzol, and blue-fluorescent 4′,6-diamidino-2-phenylindole (DAPI) nucleic acid stain were from Invitrogen (Carlsbad, California). Dihydroethidium (DHE) was from Molecular Probes (Eugene, Oregon). Supplements (Insulin solution, 3-isobutyl-1-methylxanthine (IBMX), dexamethasone, rosiglitazone), sodium arsenite, and oil red O (ORO) were from Sigma (St Louis, Missouri). Agarose ChIP kit, classic IP kit, and super signal West Pico chemiluminescent substrate were from Pierce Biochemicals (Rockford, Illinois). 5,5',6,6'-tetrachloro-1,1',3,3'-tetraethyl benzimidazolyl carbocyanine iodide (JC-1) and deoxy-2-[7-nitro-2,1,3-benzoxadiazol-4-yl) amino]-d-glucose (2-NBDG) were from Cayman chemicals (Ann Arbor, Michigan). pcDNA3.1/Sirt3 plasmid was kindly provided by Dr. Qiang Tong (Bayler College of Medicine). EGFP/SOD2 plasmid was from Addgene (Cambridge, Massachusetts). Antibodies against Sirt3 and PGC-1α were from Abcam (Cambridge, Massachusetts). Antibodies against acetylated lysine and FOXO3a were from Cell Signaling Technology (Danvers, Massachusetts). Antibody against MnSOD was from EMD Millipore (Billerica, Massachusetts). GAPDH antibody was from Santa Cruz (San Diego, California). Protease inhibitors were from Roche (Basel, Switzerland).

Cell culture and differentiation assay

Adipose and muscular tissues have the most potency for glucose expenditure. These tissues are the main targets of insulin where it regulates Glut4 metabolism, gene induction, cell trafficking and movement (Cline, et al., 1999; Martin, et al., 1999; Shepherd and Kahn, 1999). The preadipose cell line 3T3L1 was originally developed by clonal expansion from murine Swiss 3T3 cells (Green and Kehindle, 1977). Because of its potential to differentiate from fibroblasts to adipocytes, the cell line has widely been used in the studies of adipogenesis and the biochemistry of adipocytes. C2C12 myotubes are well-described insulin-responsive cell culture models for the analysis of insulin action and the development of insulin-sensitive glucose uptake mechanisms (Xue et al., 2011). 3T3L1 preadipocytes and C2C12 myotubes were obtained from ATCC (Manassas, Virginia). Cells were maintained in DMEM growth media supplemented with 10% FBS and antibiotics. For myogenic differentiation, C2C12 cells were grown to 100% confluence with 10% FBS serum and then changed to 2% FBS. By day 10, cells were fused into multinucleated myotubes. In adipogenic differentiation, 3T3L1 preadipocytes maintained in DMEM with antibiotics and 10% FBS. 3T3L1 cells were differentiated 1 day after the cells became 100% confluent (designated as day 0) using the DMI protocol (1 μM dexamethasone, 0.5 mM 3-isobutylmethylxanthine, and 1μg/mL insulin in DMEM with 10% FBS) as described previously (Yen et al., 2010). The cells were maintained at 37°C in a 5% CO₂ environment. Differentiation of preadipocytes to mature adipocytes was confirmed by ORO staining of lipid vesicles (Fu et al., 2010).

Chronic sodium arsenite exposure

To simulate more accurately human environmental exposure conditions, 3T3L1 cells and C2C12 myoblasts were exposed to low concentrations of arsenic at 0.5, 1, and 2 µM up to 8 weeks. At the end of exposure time, cells were allowed to differentiate and incubated with insulin (1 µM/ml in DMEM at 37°C, 30 min) followed by immediate sample collection for experiments as designated.

Insulin-stimulated glucose uptake

ISGU was measured in both differentiated 3T3L1 adipocytes and C2C12 myotubes using Cayman’s glucose uptake cell-based assay kit. Briefly, after treatment with arsenic, the cells were grown in six-well plates and starved for 1 h followed by incubation in 1 ml glucose-free DMEM with 100 nM insulin or vehicle (medium) up to 20 min. Glucose uptake was determined by accumulation of 2-NBDG, a fluorescently labeled deoxyglucose analog. The final concentration of 2-NBDG in the culture medium is optimized to 150 µg/ml. At the end of incubation time, cells were trypsinized and resuspended in PBS for immediate detection using flow cytometry. Cells taken up 2-NBDG display fluorescence with excitation and emission at 485 and 535 nm, respectively. Basal glucose uptake was determined without insulin.

Measurement of intracellular superoxide anion radical (O₂^•−) generation

The generation of O₂^•− was determined using a fluorescein-labeled dye DHE. Both 3T3L1 and C2C12 cells were treated with 2 µM of arsenic for 8 weeks. Thirty minutes prior to termination, 10 µM DHE was added. The cells were harvested and quantitation of ROS generation was measured using flow cytometry. For the fluorescence images analysis, the cells were mounted on coverslips. An Olympus BX53 fluorescence microscope was used to obtain images.

Mitochondrial membrane potential

The change in Δψm was analyzed by flow cytometry using Δψm sensitive dye JC-1. Briefly, both 3T3L1 adipocytes and C2C12 myotubes were cultured in 24-well plates and treated with arsenic (2 µM, 8 weeks). Cells were harvested and resuspended in 0.5 ml of cell culture media containing 10 μM JC-1 followed by measurement using FACS Calibur (BD Biosciences). Two light-scattering parameters, the forward scatter (cell size) and side scatter (intracellular granularity and complexity), and 2 fluorescence parameters, fluorescence channel 1 (FL1) and fluorescence channel 2 (FL2), were measured.

Plasmid Constructs and Transfection

To generate stable cells overexpressing mouse Sirt3 gene, pcDNA3.1/Sirt3 was transfected in C2C12 cells as described previously (Shi et al., 2005). Similarly, C2C12 stable cells expressing MnSOD gene was also established using EGFP/SOD2 plasmid. Briefly, C2C12 cells were seeded in six-well culture plates; when approximately 50% confluence, cells were transfected with 4 μg plasmid. G418 (1 mg/ml) was used to select expressing cells. Cell clones resistant to G418 were isolated. Expression of Sirt3 or MnSOD was confirmed by immunoblotting.

Immunofluorescence analysis

For detection of distribution and expression level of Sirt3, 3T3L1 cells chronically exposed to 2 µM arsenic for 8 weeks were grown on coverslips in six-well plates and allowed to differentiate as mentioned above and incubated with 400 nM MitoTracker red for 40 min. The unbound MitoTracker was removed by washing with PBS. The cells were fixed in 4% paraformaldehyde followed by permeabilization with 0.2% Triton X-100, blockage with 1 % bovine serum albumin, and incubation with Sirt3 antibody for overnight. The cells were then incubated with secondary antibody and mounted using DAPI. The cells were visualized using digital confocal microscopy (Confocal Fluorescence Imaging Microscope, Leica TCS-SP5). Similarly, immunostaining was repeated for localization of Glut4, PGC-1α, and FOXO3a. Image acquisition was carried out using Olympus BX53 fluorescence microscope.

Chromatin immunoprecipitation assay

Chromatin immunoprecipitation (ChIP) assay was performed based on the protocol provided by the manufacturer with modifications (Son et al., 2014). Briefly the cells were cross-linked by 1% formaldehyde followed by glycine treatment to stop the cross-linking. Then the cells were lysed and nuclei were digested using micrococcal nuclease. Sheared chromatin was diluted and immunoprecipitated with 2 µg of anti-FOXO3a or control IgG antibody. DNA–protein complexes were eluted from the protein A/G-agarose beads using a spin column and were reverse cross-linked by incubating with NaCl at 65°C. The relative FOXO3a binding to the PGC-1α and MnSOD promoters were analyzed by the MyiQ real-time PCR detection system (Bio-Rad) with SYBR Green PCR master mix using following primer sequences, MnSOD: (F) TTATGGAAACATTTGATAGCCACTGCTTCTTAGAC and (R) CGCGTGCTTGCTACAGCCACGC; PGC-1α: (F) CCCTGTGTACTCTTGAGATTTT and (R) CACCCTTTACTGATGGCCTTT. General PCR amplification was also performed with immunoprecipitated DNA elutes by using Mastercycler Thermal Cyclers (Eppendorf, Foster City, California).

Immunoprecipitation

Immunoprecipitation (IP) was conducted according to manufacture instruction and previous studies (Jacobs et al., 2008; Tseng et al., 2013). Differentiated cells were stimulated with insulin (100 nM/l) and then lysed using IP lysis buffer in the presence of protease inhibitors, followed by centrifugation at 14 000g for 10 min. Protein concentration of supernatant was determined and protein samples were precleared using protein A/G plus-agarose resin and incubated with primary antibodies at 4°C for 1–3 h. Fresh protein A/G plus-agarose resin was added to the lysates for overnight. Finally, the resins were washed with the lysis buffer, and the proteins were eluted using the sample loading buffer.

Immunoblotting analysis

At the end of arsenic treatment, cells were allowed to differentiate and stimulated with insulin (100 nM) at 37°C for 15 min. The whole cell lysates were extracted using lysis buffer. Thirty micrograms of protein was separated by SDS-PAGE, and incubated with primary antibodies. The blots were then reprobed with second antibodies conjugated to horseradish peroxidase. Immunoreactive bands were detected by the enhanced chemiluminescence reagent (Amersham). For the quantitation of blotting bands, the blots were exposed to Hyperfilm (Amersham Biosciences) and intensity of bands were quantified using ImageJ Densitometry software (National Institute of Health, Bethesda, Maryland).

Statistics

All values are mean ± SD of triplicates in an independent experiment, which was repeated for 3 times with the same results, and P values for comparison were calculated by student t test.

RESULTS

Arsenic Reduces ISGU in Mouse Adipocytes and Myotubes

Insulin binding to its receptor increases glucose uptake in muscle and fat, triggering a network of signaling pathways which finally stimulate the translocation of the glucose transporter Glut4 from intracellular sites to the cell membrane (Saltiel and Kahn, 2001). Previous studies have suggested that disruption of ISGU was a potential mechanism responsible for the development of T2DM in response to chronic exposures to arsenic (Paul et al., 2007). Arsenic has been demonstrated to attenuate ISGU in 3T3L1 adipocytes (Walton et al., 2004) due to an interference with the phosphoinositide-dependent protein kinase (PDK)-catalyzed phosphorylation of Akt (Paul et al., 2007). Glucose uptake measured by flow cytometry is referred to the uptake of 1 glucose analog, 2-NBDG. Our results show that exposure of 3T3L1 and C2C12 cells to arsenic caused a dose-dependent reduction in glucose uptake in response to insulin stimulation (Figs. 1A and 1B). There were no observable changes in the basal glucose uptake (without insulin stimulation) in the presence and absence of arsenic (data not shown). We then analyzed whether the reduced glucose uptake by arsenic is due, at least in part, to the decrease of Glut4 expression. Concurrently, a substantial reduction of Glut4 expression was observed in the 3T3L1 adipocytes exposed to arsenic at 1 and 2 µM (Fig. 1C). While in C2C12 myotubes, 2 µM of arsenic exposure caused a marked reduction in Glut4 expression with a comparable decrease at 1 µM of arsenic exposure (Fig. 1C). In order to determine the localization of Glut4, immunofluorescence staining in the differentiated 3T3L1 adipocytes and C2C12 cells was employed (Fig. 1D). The results show that untreated 3T3L1 cells exhibited a high expression of Glut4 in perinuclear region, while arsenic-treated cells displayed a markedly decrease of Glut4 expression in the same region (Fig. 1D, top). However, Glut4 was located in the nucleus and surrounding region in untreated C2C12 cells. Upon arsenic treatment for 8 weeks, Glut4 expression was reduced both in the nucleus and its surrounding region compared to control without arsenic treatment (Fig. 1D, bottom).

Arsenic Causes O₂^•− Generation and Decreases Δψm

The ability of arsenic to induce oxidative stress in exposed cells is well established. Arsenic has been demonstrated to generate ROS, leading to oxidative damage to DNA, lipids, and proteins (Jomova and Valko, 2011; Schwerdtle et al., 2003). In this study, the ability of arsenic to generate O₂^•− was quantified by flow cytometry using fluorescent dye DHE. O₂^•− generation was dramatically elevated (2-fold) in both 3T3L1 adipocytes and C2C12 myotubes treated with 2 µM of arsenic for 8 weeks (Fig. 2A). To investigate whether arsenic-induced ROS generation causes mitochondrial damage, Δψm analysis was performed. Carbonyl cyanide m-chlorophenyl hydrazone (CCCP) was used as a positive control which disrupts Δψm. The result shows that arsenic treatment significantly decreased Δψm in both 3T3L1 adipocytes and C2C12 myotubes (Fig. 2B), indicating that chronic arsenic exposure caused mitochondrial damage. To determine whether arsenic-decreased ISGU expression is regulated through mitochondrial antioxidant mechanism, both 3T3L1 cells and C2C12 cells with stable overexpression of MnSOD gene were established. This manipulatively increased expression of MnSOD led to a substantial enhancement in the recovery of reduced ISGU in the arsenic-treated 3T3L1 cells and C2C12 cells (Fig. 2C), suggesting that mitochondrial damage is involved in the reduced glucose uptake by arsenic.

Arsenic Decreases Sirt3 Expression in Mouse Adipocytes and Myotubes

Sirt3 is a mitochondrial oxidative stress response protein. To investigate involvement of Sirt3 in arsenic-induced ISGU reduction, Sirt3 expression was examined in both fully differentiated 3T3-L1 and C2C12 cells. The results from immunoblotting analysis show that Sirt3 expression was decreased as early as 2 weeks of arsenic treatment in both 3T3L1 cells and C2C12 cells in a dose- and time-dependent manner (Fig. 3A). The intensity of those blotting bands were quantified and the results confirmed the observations of Figures 3A and 3B. Results from immunostaining show that in the 3T3-L1 cells without arsenic treatment, Sirt3 was present in perinuclear region (Fig. 3C, left). By using MitoTracker red dye, Sirt 3 was identified to locate in the inner mitochondrial membrane (Fig. 3D, top). Sirt3 was translocated from mitochondria to nucleus at 2 weeks of arsenic treatment (Fig. 3D, top). Its expression was almost diminished at 8 weeks of arsenic treatment (Figs. 3D, top). However, in the untreated C2C12 cells Sirt3 was located both in the nucleus and mitochondria (Fig. 3C, left and Fig. 3D, bottom). Upon 8 weeks of arsenic treatment, expression of Sirt3 in both mitochondria and nucleus was dramatically reduced (Fig. 3C, left and Fig. 3D, bottom).

Arsenic Suppresses Sirt3 Target Proteins FOXO3a, MnSOD, and PGC-1α

To explore whether downstream target proteins of Sirt3 were involved upon chronic arsenic treatment, expression of FOXO3a was analyzed using immunoblotting and fluorescence immunostaining. The result shows that treatment of 3T3L1 and C2C12 cells with arsenic up to 8 weeks caused increased phosphorylation of FOXO3a at serine 253, an inactivated form of FOXO3a, in a dose-dependent manner (Fig. 4A). The increased FOXO3a phosphorylation triggered nuclear export to cytoplasm, leading to its inactivation in differentiated mouse adipocytes as shown by immunofluorescence analysis (Fig. 4B). It has previously been suggested that nuclear localization of FOXO3a is essential for its transcriptional activity and is precisely controlled by multiple post-translational modifications (Calnan and Brunet, 2008; Vogt et al., 2005). We then examined deacetylation of FOXO3a by Sirt3. The result shows that acetylation of FOXO3a residue at K-100 lysine was markedly increased in the arsenic-exposed C2C12 myoblasts (Fig. 4C). Next, physical interaction of Sirt3 with the FOXO3a was examined by IP analysis. The result shows that binding of Sirt3 to FOXO3a was dramatically reduced in chronic arsenic-exposed cells compared to their respective control (Fig. 4D).

To further confirm that reduced expressions MnSOD and PGC-1α by arsenic exposure are due to decreased bindings of FOXO3a, ChIP analysis was performed. The results show that binding of FOXO3a to MnSOD or PGC-1α gene promoter was prominently decreased in arsenic-treated cells compared respective control C2C12 cells (Fig. 4E). Quantification analysis show that binding of FOXO3a to either MnSOD or PGC-1α was significantly reduced in arsenic-treated cells compared to control ones without treatment (Fig. 4F), indicating that MnSOD and PGC-1α are direct targets of FOXO3a.

Sirt3 Restores ISGU Inhibited by Arsenic Through MnSOD

Next, we examined whether Sirt3 overexpression could restore C2C12 cells from arsenic-induced inhibition of ISGU. For this purpose, Sirt3 expression vector was stably transfected in C2C12 cells (Fig. 5A). Overexpression of Sirt3 partially restored ISGU reduced by arsenic exposure, concomitantly with increased expression of Glut4 (Figs. 5B and 5C). Manipulated Sirt3 expression restored activation of FOXO3a by decreased phosphorylation at Ser253 and expressions of both MnSOD and PGC-1α reduced by arsenic exposure (Fig. 5C).

It has previously been reported that Sirt3-mediated deacetylation of FOXO3a leads to upregulation of FOXO3a which is dependent on mitochondrial antioxidant enzymes (Sundaresan et al., 2009). Our results show that overexpression of Sirt3 caused a comparable increase of FOXO3a deacetylation (Fig.5D), leading to upregulation of FOXO3a-dependent mitochondrial antioxidant enzymes, PGC-1α and MnSOD which facilitate ROS detoxification in response to chronic arsenic exposure (Fig. 5C). Taking together, the results suggest that enhanced antioxidants protect cells from mitochondrial damage, resulting in restoration of ISGU decreased by chronic arsenic exposure.

To study whether manipulated expression of Sirt3 is able to restore binding of FOXO3a to MnSOD or PGC-1α, ChIP analysis was performed. The results show that overexpression of Sirt3 restored reduced binding of FOXO3a to gene promoter of either MnSOD or PGC-1α induced by arsenic treatment (Figs. 5E and 5F), suggesting that Sirt3 enhances the transcription of FOXO3a-dependent antioxidant genes by promoting the nuclear localization of FOXO3a (Sundaresan et al., 2009). MnSOD is a mitochondrial antioxidant; therefore, mitochondrial potential was evaluated in the Sirt3 expressed cells. The results show that overexpression of Sirt3 inhibited the disruption of mitochondrial potential induced by arsenic exposure (Fig. 5G). MnSOD overexpressed cells exhibited potential capability in regaining the loss of mitochondrial potential induced by arsenic compared to Sirt3 overexpressed ones (Fig. 5G), implicating that Sirt3/MnSOD is a positive cofactor in regaining mitochondrial damage induced by arsenic exposure. Overexpression of Sirt3 was able to inhibit ROS generation (Fig. 5H), and this may contributes to regain the reduced Δψm by arsenic exposure.

DISCUSSION

Oxidative stress from excessive ROS production and mitochondrial dysfunction are the major factors in the development of insulin resistance in T2DM. Exposure to differentiated 3T3L1 cells to low levels of arsenic leads to an increased oxidative stress environment, a decreased ISGU, and a reduction in Glut4, a glucose transporter critical to fat and muscle uptake of glucose, contributing to arsenic-induced insulin resistance (Xue et al., 2011). The link between arsenic and diabetes is well described through human and animal studies (Paul et al., 2011). Increased levels of oxidative stress have been demonstrated in skeletal muscle of type 2 diabetic mice (Yokota et al., 2009) and in patients with diabetes (Ramakrishna and Jailkhani, 2008). Sirt3 expression in skeletal muscle is significantly decreased in rodent models of type 1 and type 2 diabetes and is regulated by fasting and caloric restriction (CR) (Jing et al., 2011). It is evident from a recent report that islet expression of mitochondrial Sirt3 mRNA is decreased in type 2 diabetic humans (Caton et al., 2013).

Given that mitochondria impairments and inhibition in ISGU are central to T2DM and that Sirt3 promotes mitochondrial health, the present study investigates the molecular mechanism in which how both excess ROS production and mitochondrial dysfunction further stimulate various signaling pathways, leading to reduced ISGU in response to chronic arsenic exposure. We revealed that chronic arsenic exposure induced oxidative stress may be a pivotal factor for a decrease in the mitochondrial membrane potential, resulting in prevention of antioxidants from their ability to inhibit ISGU and thus insulin resistance by scavenging free radicals. Antioxidant administration has been shown to decrease manifestations of insulin resistance in mice (Hildebrandt et al., 2004), indicating the role of ROS in inhibiting insulin action. Low level of Sirt3 is associated with increased cellular ROS levels (Bell et al., 2011; Caton et al., 2013; Jing et al., 2011, 2013; Sundaresan et al., 2009). Skeletal muscles in Sirt3 knockout mice exhibited an alternated mitochondrial function and increased levels of oxidative stress (Jing et al., 2011). Increased ROS level in Sirt3 knockdown C2C12 cells impaired insulin signaling (Jing et al. 2011). As expected, we have shown that Sirt3 knockdown led to increased cellular ROS levels in C2C12 cells (Supplementary Data).

Recently, Sirt3 has gained increasing attention as a regulator of metabolic activity in muscle (Joseph et al., 2011). Sirt3 contains a cleavable N-terminal mitochondrial targeting signal that permits its import into mitochondrial subcompartments (Chen and Guarente, 2007). Sirt3 levels were diminished with aging, high fat diets, and in response to exercise, caloric restriction, and fasting (Lanza et al., 2008; Palacios et al., 2009; Someya et al., 2010). Overexpression of Sirt3 showed its antioxidant regulatory effect in primary cardiomyocytes, suggesting that Sirt3 may regulate the antioxidant systems (Sundaresan et al., 2009). Consistently, our study showed that Sirt3 overexpression reduced ROS generation with or without arsenic exposure. Previous study has shown a link between Sirt3 and mitochondrial ROS production by targeting MnSOD under different pathological and physiological conditions (Bell et al., 2011). Overexpression of Sirt3 in brown adipocytes led to increased expression of PGC-1α and decreased intracellular ROS level (Shi et al., 2005). Another study has demonstrated that PGC-1α suppressed ROS production through induction of ROS-detoxifying enzymes (St-Pierre et al., 2006). This study has demonstrated that both Sirt3 and MnSOD are cytoprotective in cultured mouse myotubes since they stabilize mitochondrial membrane potential, reduce ROS generation, and prevent reduced ISGU by chronic arsenic exposure, suggesting a link between chronic arsenic-induced oxidative stress and resultant mitochondrial depolarization that may prevent antioxidants from their ability to inhibit arsenic reduced ISGU. Sirt3 can interact with FOXO3a in the nucleus and in mitochondria (Jacobs et al., 2008; Sundaresan et al., 2009). Using confocal microscope analysis, this study has shown that Sirt3 is localized in the inner mitochondrial membrane in 3T3L1 adipocytes. While FOXO3a and PGC-1α were preferentially seen in the nucleus of the control cells without arsenic treatment, as evidenced from our immunofluorescence study. These findings raise the possibility that in 3T3L1 adipocytes, Sirt3 deacetylates FOXO3a in the nucleus to build the nucleus-mitochondria communications for diverse regulation of mitochondria (Tseng et al., 2013). Conversely, Sirt3 was shown to be located both in the nucleus and mitochondria in C2C12 myotubes. This may be due to a retrograde communication through Sirt3-FOXO3a in the nucleus (Tseng et al., 2013). Mitochondrial ROS can be reduced by Sirt3-FOXO3a through an increase in MnSOD (Sundaresan et al., 2009) or through upregulation of PGC-1α to interact with FOXO3a transcriptional activation of a set of mitochondrial antioxidant enzymes (Olmos et al., 2009). It has been reported that deacetylation of FOXO3a by Sirt3 promotes the translocation of FOXO3a from mitochondria to nucleus, resulting in a complex influence on the transcription of nuclear genes controlling mitochondria (Tseng et al., 2013). Sirt3 targets PGC-1α and influences mitochondrial transcriptional regulation in muscle (Kong et al., 2010; Palacios et al., 2009). PGC-1α is one of the most well-known regulators of mitochondrial biogenesis (Scarpulla, 1997; Wu et al., 1999). In particular, PGC-1α binds and coactivates transcription factors such as the estrogen-related receptor alpha (ERRα) and the nuclear respiratory factors 1 and 2 (NRF-1 and NRF-2) to cause induction of genes involved in metabolism and mitochondrial biogenesis (Wu et al., 1999). Animals with increased muscle PGC-1α have a longer lifespan that is associated with enhanced mitochondrial function, improved insulin sensitivity, and reduced oxidative damage (Wenz et al., 2009). PGC-1α mRNA levels are reduced in certain cohorts of obese and type 2 diabetic individuals (Mootha et al., 2003; Patti et al., 2003). In some populations, polymorphisms in the PGC-1α gene have been linked to a predisposition for type 2 diabetes (Ek et al., 2001; Kim et al., 2005). Previous studies in the molecular mechanism of PGC-1α reveal that PGC-1α induces GLUT4 expression by interacting and coactivating myocyte-enhanced factor 2 (MEF2) transcription regulator (Michael et al., 2001), indicating that PGC-1α is a central messenger of nuclear-mitochondrial crosstalk during cellular stress (Joseph et al., 2011).

Sirt3 has been reported to alter the acetylation status of a number of proteins linked to ROS production (Ahn et al., 2008; Jing et al., 2011; Lombard et al., 2007; Qiu et al., 2010; Schlicker et al., 2008; Tseng et al., 2013; Yu et al., 2012). This study has demonstrated that Sirt3 interacts with and deacetylates FOXO3a, a FOXO transcription factor, to induce its antioxidant effects. Deacetylation of FOXO3a by Sirt3 further reduces its phosphorylation at Ser253. Recent study has suggested that arsenic could induce phosphorylation of FOXO transcription factors (Hamann et al., 2014). Our ChIP study shows that Sirt3 overexpression stimulated expression of FOXO target genes, MnSOD and PGC-1α, leading to reduced ROS generation. The most important and novel finding of this study is that enforced Sirt3 expression in C2C12 cells under chronic arsenic conditions recovered arsenic inhibited ISGU. Meanwhile, knockdown of Sirt3 by shRNA induced an inhibition in ISGU. The present study has found that lysine residue (Ac-100) of FOXO3a was hyperacetylated by Sirt3 in chronic arsenic-treated C2C12 cells, suggesting that FOXO3a may also be the direct target of Sirt3. Overexpression of Sirt3 in C2C12 cells showed a decrease in acetylation and this decrease was almost similar when these cells were subjected to chronic arsenic exposure.

In summary, the decreased level of Sirt3 in mouse skeletal muscle and adipocytes upon chronic arsenic exposure is an important factor in the inhibition of ISGU. This decrease induced ROS generation and oxidative stress and subsequently caused mitochondrial membrane damage, leading to insulin resistance. This study provides direct evidence of the importance of reversible acetylation/deacetylation in the mitochondria/nucleus and its potential role in the development of insulin resistance. We postulate that pharmaceutical interventions that induce Sirt3 activity or expression by increasing oxidative stress resistance or prevent mitochondrial dysfunction or both could potentially reduce arsenic induced T2DM.

ACKNOWLEDGMENT

This study was supported by University of Kentucky to Zhang Z.

REFERENCES

Author notes