Sirtuin3 Dysfunction Is the Key Determinant of Skeletal Muscle Insulin Resistance by Angiotensin II

Background Angiotensin II promotes insulin resistance. The mechanism underlying this abnormality, however, is still poorly defined. In a different setting, skeletal muscle metabolism and insulin signaling are regulated by Sirtuin3. Objective Here, we investigate whether angiotensin II-induced insulin resistance in skeletal muscle is associated with Sirtuin3 dysregulation and whether pharmacological manipulation of Sirtuin3 confers protection. Study Design Parental and GLUT4-myc L6 rat skeletal muscle cells exposed to angiotensin II are used as in vitro models of insulin resistance. GLUT4 translocation, glucose uptake, intracellular molecular signals such as mitochondrial reactive oxygen species, Sirtuin3 protein expression and activity, along with its downstream targets and upstream regulators, are analyzed both in the absence and presence of acetyl-L-carnitine. The role of Sirtuin3 in GLUT4 translocation and intracellular molecular signaling is also studied in Sirtuin3-silenced as well as over-expressing cells. Results Angiotensin II promotes insulin resistance in skeletal muscle cells via mitochondrial oxidative stress, resulting in a two-fold increase in superoxide generation. In this context, reactive oxygen species open the mitochondrial permeability transition pore and significantly lower Sirtuin3 levels and activity impairing the cell antioxidant defense. Angiotensin II-induced Sirtuin3 dysfunction leads to the impairment of AMP-activated protein kinase/nicotinamide phosphoribosyltransferase signaling. Acetyl-L-carnitine, by lowering angiotensin II-induced mitochondrial superoxide formation, prevents Sirtuin3 dysfunction. This phenomenon implies the restoration of manganese superoxide dismutase antioxidant activity and AMP-activated protein kinase activation. Acetyl-L-carnitine protection is abrogated by specific Sirtuin3 siRNA. Conclusions Our data demonstrate that angiotensin II-induced insulin resistance fosters mitochondrial superoxide generation, in turn leading to Sirtuin3 dysfunction. The present results also highlight Sirtuin3 as a therapeutic target for the insulin-sensitizing effects of acetyl-L-carnitine.


Objective
Here, we investigate whether angiotensin II-induced insulin resistance in skeletal muscle is associated with Sirtuin3 dysregulation and whether pharmacological manipulation of Sir-tuin3 confers protection.

Study Design
Parental and GLUT4-myc L6 rat skeletal muscle cells exposed to angiotensin II are used as in vitro models of insulin resistance. GLUT4 translocation, glucose uptake, intracellular molecular signals such as mitochondrial reactive oxygen species, Sirtuin3 protein expression and activity, along with its downstream targets and upstream regulators, are analyzed both in the absence and presence of acetyl-L-carnitine. The role of Sirtuin3 in GLUT4 translocation and intracellular molecular signaling is also studied in Sirtuin3-silenced as well as overexpressing cells.

Results
Angiotensin II promotes insulin resistance in skeletal muscle cells via mitochondrial oxidative stress, resulting in a two-fold increase in superoxide generation. In this context, reactive oxygen species open the mitochondrial permeability transition pore and significantly lower Sirtuin3 levels and activity impairing the cell antioxidant defense. Angiotensin II-induced Sir-tuin3 dysfunction leads to the impairment of AMP-activated protein kinase/nicotinamide Introduction Insulin resistance is a key component of metabolic syndrome, a cluster of hypertension, diabetes mellitus, dyslipidemia, and chronic inflammation, observed frequently in subjects who are obese or have a family history of diabetes, and renal or cardiovascular disease [1,2]. Metabolic syndrome affects approximately 30% of adults in Europe [3], but this percentage is rapidly increasing worldwide, particularly in emerging countries, indicating there may be a diabetes epidemic in the next few decades [4]. Identifying the mechanistic origin of insulin resistance remains a major objective in designing innovative treatments.
One of several insulin-sensitive tissues, skeletal muscle plays a well-established role in maintaining glucose homeostasis and in regulating whole body carbohydrate metabolism. Skeletal muscle is the major site of insulin-mediated glucose uptake [5,6] and insulin resistance in this tissue causes metabolic syndrome [7] and is recognized as the initial metabolic defect for the development of type 2 diabetes [8].
Multiple factors contribute to insulin resistance, involving interacting mechanisms such as ectopic lipid accumulation, endoplasmic reticulum stress, pro-inflammatory response, and, importantly, the activation of the renin-angiotensin system (RAS) [9,10]. The major effector of RAS is angiotensin II (Ang II)-a critical promoter of insulin resistance because of its effect on insulin receptors and downstream signaling, which results in desensitization to insulin in metabolic tissues [11]. The intracellular mechanism through which Ang II impairs insulin signaling in skeletal muscle cells and in isolated skeletal muscles is thought to be NADPH-oxidase-derived reactive oxygen species (ROS) [12][13][14]. Based on the evidence that insulin resistance is associated with mitochondrial dysfunction [15][16][17], researchers have increasingly focused on the possibility that insulin resistance and metabolic syndrome are dependent on altered mitochondrial oxidation and reactive oxygen species production [16,18]. Chronic Ang II infusion causes glucose intolerance and mitochondrial abnormalities in skeletal muscles of healthy mice together with increased mitochondrial ROS generation [19]. In a model of insulin resistance associated with excessive RAS activation, mitochondrial abnormalities are observed in the liver, skeletal muscles and the myocardium [15]. Whether or not angiotensin II-induced insulin resistance is causally linked to mitochondrial ROS generation and the downstream target(s) of this phenomenon remains to be ascertained.
We previously described a link between Ang II and Sirtuin3 (Sirt3), the primary mitochondrial NAD + -dependent deacetylase that regulates energy and redox homeostasis, protecting mitochondria from oxidative damage [20]. Mice deficient in the Agtr1 gene-which encodes the Ang II type 1a receptor-showed multiple organ protection from age-induced oxidative stress and preservation of mitochondrial numbers in renal tubules linked with Sirt3 upregulation [21].
Modulators of mitochondrial metabolism have been studied in an attempt to ameliorate metabolic abnormalities associated with mitochondrial dysfunction in diverse experimental and human conditions. One of these is the recent finding that acetyl-L-carnitine (ALCAR) [22] ameliorated glucose tolerance and insulin resistance in subjects with a cluster of risk factors for diabetes mellitus and cardiovascular disease [23].
In the present study, we investigate whether angiotensin II-induced insulin resistance in skeletal muscle is associated with Sirt3 dysregulation and whether pharmacological manipulation of Sirt3 confers protection.

2-deoxyglucose uptake
The assay was performed in L6 myotubes as described by Yonemitsu et al. [26]. Specific 2-deoxyglucose uptake was expressed as pmol/min/mg protein from determination performed at least in triplicate.

Subcellular fractionation
The subcellular fractionation of L6 myotubes was performed as described by Mitsumoto and Klip [27] with slight modifications. After incubations, cells were gently scraped and incubated in hypotonic lysis buffer (10 mM Tris-HCl pH 7.4, 2 mM EDTA, 200 μM PMSF, 1 mM benzamidine, 10 μg/mL pepstatin and 10 μg/mL leupeptin) for 20 min on ice and then lysed by sonication. An aliquot of the total cell lysate was saved for western blot analysis of the total GLUT4 and the remaining sample was centrifuged at 1,000 g for 10 min at 4°C to remove nuclei and unbroken cells. The supernatant was centrifuged at 31,000 g for 60 min to pellet crude plasma membrane (CPM).

Measurement of GLUT4-myc translocation
Detection of GLUT4-myc on the cell surface of intact L6 GLUT4-myc myotubes was assessed using a colorimetric-based assay [24].

Immunocytochemistry
Paraformaldehyde-fixed cells were incubated with Triton 0.3% and, after blocking, with a rabbit monoclonal anti-Sirt3 antibody (1:100, Cell Signaling Technology Inc., Danvers, MA), followed by a goat anti-rabbit Cy3-conjugated secondary antibody (Jackson ImmunoResearch Laboratories, Baltimore Pike, PA). Nuclei were stained with DAPI (Sigma-Aldrich). Sirt3 staining was examined under confocal laser scanning microscopy (LS 510 Meta, Zeiss). Sirt3 protein expression was quantified in 10 randomly acquired fields per sample. Specifically, the area corresponding to the Sirt3 staining was measured in pixels 2 using Image J 1.40g software and normalized for the number of DAPI-positive nuclei.

Western blot analysis
Whole cells, subcellular fractions or isolated mitochondria from L6 myotubes were analyzed. Proteins were separated on SDS-PAGE under reducing conditions and transferred to PVDF membranes (Bio-Rad Laboratories) that were incubated overnight at 4°C with primary antibodies. The following antibodies were used: mouse monoclonal anti-GLUT4 antibody (

SOD activity assay
SOD activity in isolated mitochondria was measured using a Colorimetric Activity kit (Arbor Assays, Ann Arbor, MI).
Sirt3 silencing L6 GLUT4-myc and parental L6 myotubes were transfected with ON-TARGET plus SMART pool Sirt3 (100 pmol) duplex for the target sequence NM_001106313, or with control nontarget siRNA (Ambion, Silencer Select Negative Control #2siRNA) using Lipofectamine 2000 reagent (Invitrogen) according to the manufacturer's protocol. Forty-eight h after transfection, L6 GLUT4-myc myotubes were used for GLUT4 translocation assessment. Western blot analysis of pAMPK/AMPK was performed in L6 cells 96 h after transfection.

Sirt3 overexpression
L6 myotubes were transfected with plasmid DNA (2,5 μg GFP-tagged pCMV-hSirt3, Origene Technologies) using Lipofectamine 2000 (Invitrogen, Life Technologies) as described in the manufacturer's instructions. Afterwards cells were exposed to medium with or without Ang II for 15 h.

Statistical analysis
Results are expressed as mean ± SE. Data analysis was performed using the computer software Prism (GraphPad Software, Inc. San Diego USA). Comparisons were made by ANOVA with the Bonferroni post hoc test or unpaired Student's t test as appropriate. Statistical significance was defined as P < 0.05.

Results
Ang II promotes insulin resistance via mitochondrial ROS, which is reverted by ALCAR Ang II triggered mitochondrial ROS generation in L6 myotubes as documented by a two-fold increase over controls both in the cellular production of O 2 • (MitoSOX, mean fluorescence intensity-MFI-) and in the percentage of cells undergoing oxidative stress (Fig 1A). This effect was associated with a 59% inhibition of insulin-stimulated GLUT4 translocation to the cell plasma membrane (Fig 1B) and the reduction of cell glucose uptake (Fig 1C). Treatment with ALCAR significantly reduced excessive mitochondrial O 2 • and prevented Ang II-induced insulin resistance (Fig 1A-1C). To further confirm results obtained in parental L6 myotubes, L6 muscle cells stably expressing GLUT4 tagged on an extracellular domain with a myc epitope were used. Ang II also impaired insulin-stimulated GLUT4 transport in GLUT4-myc L6 myotubes, which was restored by ALCAR (Fig 1D). Similarly, MnTBAP, a mitochondriapenetrating SOD mimetic, rescued GLUT4-myc cells from Ang II-induced insulin resistance confirming that scavenging mitochondrial O 2 • improved insulin sensitivity (Fig 1D). ALCAR alone did not affect the insulin sensitivity of L6 myotubes evaluated as GLUT4 translocation (cell surface GLUT4-myc fold changes, ALCAR + Ins 3.08 ± 0.23 vs. Ins 3.47 ± 0.06). Notably, Ang II alone significantly reduced apical expression of GLUT4 in GLUT4-myc L6 myotubes, an effect that was prevented by ALCAR (S1 Fig). Ang II via ROS inhibits Sirt3 and impairs mitochondrial antioxidant defense, both of which are restored by ALCAR In physiological conditions the accumulation of ROS within mitochondria is counteracted by antioxidant enzymes whose activity is regulated by the NAD + -dependent mitochondrial deacetylase Sirt3 [20]. Ang II markedly reduced Sirt3 protein expression in L6 myotubes as revealed by lower immune reactive speckled perinuclear staining compared to unstimulated cells (Fig  2A). Consistently, Sirt3 protein content in isolated mitochondria was reduced by 50% in response to Ang II, while normal levels were found in the presence of ALCAR (Fig 2B). Ang II also affected Sirt3 deacetylase activity as seen in the evaluation of protein acetylation in mitochondria from unstimulated or Ang II-treated L6 myotubes. Ang II induced hyperacetylation of mitochondrial proteins with molecular weights of approximately 55, 48, 44, 38, and 26 kDa, the latter corresponding to MnSOD (Fig 3A). This effect was concomitant with the reduction of MnSOD antioxidant activity (Fig 3B). The addition of ALCAR reversed the inhibitory effect of Ang II on Sirt3 activity (Fig 3A, left) and restored the cell antioxidant defense, normalizing MnSOD activity ( Fig 3B)   fostering MnSOD activity (Fig 3B), limited mitochondrial protein acetylation induced by Ang II in L6 myotubes (Fig 3A, right) confirming an inhibitory role for oxygen radicals in Sirt3-dependent mitochondrial protein deacetylation.
To assess whether impaired Sirt3 deacetylase activity induced by Ang II was due to the deleterious effect of mitochondrial O 2 • opening the mitochondrial permeability transition pore (mPTP) leading to the collapse of ΔC, we investigated the effect of Ang II on ΔC -sensing dye JC-1 in both the absence and presence of ALCAR or MnTBAP. Ang II-treated L6 myotubes showed mitochondrial depolarization, as evidenced by diffuse cytoplasmic JC-1 green staining, that was prevented by both ALCAR and MnTBAP (Fig 4A-4D

Ang II-induced Sirt3 dysfunction leads to impairment of AMPK/NAMPT signaling
To elucidate the pathways linking Sirt3 dysfunction-induced by Ang II-to impaired GLUT4 translocation, we focused on AMPK, a master regulator of cellular energy metabolism involved in glucose transport in skeletal muscles [28,29]. Ang II reduced the pAMPK/AMPK ratio by about 70% in L6 myotubes (Fig 5A, left). ALCAR normalized AMPK phosphorylation in Ang II-treated L6 myotubes (Fig 5A, left), while it did not affect the pAMPK/AMPK ratio in control cells (fold changes vs. control, ALCAR 0.99 ± 0.16 vs. 1.0). Scavenging mitochondrial O 2 • by MnTBAP yielded similar results on AMPK phosphorylation in Ang II-treated and control L6 myotubes (Fig 5A, left). Activation of AMPK by AICAR [30] normalized GLUT4-myc levels on the surface of Ang II-treated cells (Fig 5B), while inhibition of AMPK by compound C [31] mimicked the inhibitory effect of Ang II on GLUT4 translocation ( Fig 5B) suggesting a functional link between AMPK inactivation and impaired glucose transport.
To establish the role of Sirt3 in AMPK regulation we studied the effect of Ang II on the pAMPK/AMPK ratio in Sirt3 overexpressing cells (pSirt3). A 4-fold increase in Sirt3 protein expression (S4 Fig) abrogated Ang II-induced inactivation of AMPK (Fig 5A, middle). On the other hand, silencing Sirt3 with small interfering RNA (siSirt3) resulted in a 69% decrease of Sirt3 mRNA compared to scrambled cells (0.31 ± 0.05 vs. 1.0 ± 0.15, P < 0.05) and led to a significant reduction of the pAMPK/AMPK ratio (Fig 5A, right) confirming the regulatory effect of Sirt3 on AMPK.
In skeletal muscles AMPK is known to affect the expression of NAMPT, the rate-limiting enzyme in the NAD salvage pathway [32,33]. Consistent with AMPK inhibition, Ang II down-regulated NAMPT mRNA levels that were normalized by ALCAR (Fig 5B). The inhibitory effect of Ang II on AMPK was comparable to that of compound C confirming AMPK as an upstream regulator of NAMPT in L6 myotubes (Fig 5C).

ALCAR protection of skeletal muscle cells against Ang II-induced insulin resistance requires Sirt3
To further demonstrate the functional role of Sirt3 in the protective effect of ALCAR on Ang II-induced insulin resistance, we evaluated cell surface GLUT4 expression in Sirt3 siRNA and   Sirt3 Dysregulation in Angiotensin II-Induced Insulin Resistance scrambled cells (0.31 ± 0.02 vs. 1.0 ± 0.01, P < 0.0001). As shown in Fig 6, insulin promoted a 3-fold increase in cell surface GLUT4-myc in irrelevant siRNA myotubes that was significantly reduced by Ang II. ALCAR prevented the inhibitory effect of Ang II on insulin-mediated GLUT4 transport in irrelevant siRNA cells (Fig 6). Silenced Sirt3 myotubes became resistant to insulin that failed to stimulate GLUT4-myc translocation to the cell surface (Fig 6). Importantly, knocking out Sirt3 negated the effect of Ang II as well as the protective action of ALCAR on GLUT4 translocation.

Discussion
Here we found that 1) Ang II promotes insulin resistance through mitochondrial ROS, 2) enhanced mitochondrial O 2 • production leads to Sirt3 dysfunction, impairing mitochondrial antioxidant defense, 3) ALCAR protects against Ang II-induced insulin resistance by preventing Sirt3 dysfunction. The present results are consistent with data that chronic delivery of mitochondrial SOD mimetic improved glucose homeostasis in ob/ob mice [34] and that MnSOD transgenic mice were protected from high fat diet-induced insulin resistance [35]. Accumulated ROS contribute to mitochondrial dysfunction through the mPTP opening that depletes mitochondrial NAD + , the substrate for Sirt3 deacetylase activity [36]. Our findings that MnTBAP prevented Ang IIinduced mitochondria depolarization and acetylation of mitochondrial proteins would indicate that O 2 • , by opening mPTP, leads to Sirt3 dysregulation, by activating a feed-forward loop that sustains oxidative stress in skeletal muscle cells. Previous evidence in cultured renal tubular epithelial cells of a link between Ang II and Sirt3 via Ang II type 1 receptor (AT1R) [21], suggests a possible role of AT1R in Ang II-induced Sirt3 dysfunction in the present setting. Sirt3 activity can be regulated by AMPK through NAMPT, the rate-limiting enzyme in the biosynthesis of Sirt3 substrate NAD [37]. In this context, it is reported that AMPK signaling regulates NAMPT mRNA and protein expression in skeletal muscles [32,33]. Our results showing that down-regulation of NAMPT was secondary to AMPK inhibition indicate that AMPK has a causative role in modulating NAMPT gene transcription, and possibly Sirt3 deacetylase activity in response to Ang II. AMPK regulates insulin action [38][39][40] and is a drug Sirt3 Dysregulation in Angiotensin II-Induced Insulin Resistance target for diabetes and metabolic syndrome [40][41][42]. When AMPK was inhibited by Ang II, there was reduced cell surface GLUT4 expression, which was reversed by the AMPK agonist AICAR. Our findings are in line with the evidence that Ang II inhibits AMPK-dependent glucose uptake in the soleus muscle [43] and that AMPK activation is part of the protective effect of angiotensin receptor blockade against Ang II-induced insulin resistance [44].
To add to the complexity, one might consider that excessive oxygen radical production also negatively regulates AMPK function. There is already evidence that AMPK can be activated by Sirt3 when it deacetylates LKB1 [45], the primary upstream kinase of AMPK. Moreover, skeletal muscles from Sirt3-deficient mice show reduced AMPK phosphorylation [46], while Sirt3 Dysregulation in Angiotensin II-Induced Insulin Resistance increased muscle AMPK activation is observed in transgenic mice with muscle-specific expression of the murine Sirt3 short isoform [47].
Previous studies in L6 rat skeletal muscle cells showed that Ang II impairs insulin signaling by inhibiting insulin-induced tyrosine phosphorylation of insulin receptor substrate 1 (IRS-1) and the activation of Akt [12]. Similarly, Sirt3 deletion in cultured myoblasts impairs insulin signaling, leading to a decrease in tyrosine phosphorylation of IRS-1 [48]. It is conceivable that Ang II-induced Sirt3 dysfunction in our setting negatively regulates insulin metabolic signaling, affecting both IRS-1 and the distal downstream step Akt activation.
Our study focused on mitochondrial ROS as a driver of Ang II-induced insulin resistance in skeletal muscle cells. However, NADPH oxidase has been also reported as a source of ROS induced by Ang II in L6 myotubes [12]. The relative role of NADPH oxidase and mitochondria in ROS generation in Ang II-treated skeletal muscle cells is unknown. There is emerging evidence of cross talk between NADPH oxidase and mitochondria in regulating ROS generation. In different settings, NADPH oxidase-derived ROS can trigger mitochondrial ROS formation and vice-versa [49][50][51]. It is conceivable that Ang II-induced NADPH oxidase activation would concur to trigger mitochondrial changes in L6 myotubes.
Disorders characterized by mitochondrial dysfunction and oxidative stress, such as neurodegeneration and cognitive deficit [52,53], benefit from ALCAR supplementation, which seems to act as an antioxidant, likely by improving mitochondrial efficiency [54,55]. In this study, we provide evidence that by preventing the inhibitory effect of Ang II on Sirt3 expression and activity, ALCAR restored the antioxidant activity of MnSOD, rescuing skeletal muscle cells from mitochondrial superoxide-driven insulin resistance. Here the beneficial effects of ALCAR in improving insulin sensitivity vanished when Ang II-treated myotubes were silenced for Sirt3, underlining that the antioxidant effects of ALCAR depend on Sirt3-mediated Sirt3 Dysregulation in Angiotensin II-Induced Insulin Resistance mitochondria protection. Our data showing that ALCAR normalized NAMPT expression through the activation of AMPK additionally support Sirt3 as a target for the insulin-sensitizing effect of ALCAR. That the compound may act on the AMPK pathway is also suggested by previous findings in rat skeletal muscle cells [56] and soleus muscles [57].
Thus, Sirt3 could be the unifying intracellular molecular signaling through which L-carnitine and its esters, including ALCAR, protect mitochondria and ameliorate insulin resistance. This is relevant in view of the emerging role of L-carnitine and its derivatives as promising treatment for diseases associated with mitochondrial dysfunction [58]. In this context propionyl-L-carnitine has been shown to improve mitochondrial respiratory chain activity in the livers of diet-induced obese mice and to protect these animals from insulin resistance and cardiovascular complications [59].
One limitation of the present findings is that they are gathered from cultured skeletal muscle cells. In vivo follow-up studies in experimental models of RAS-related insulin resistance are needed to definitely prove the functional relevance of these findings.
In conclusion, our data clarify and explain the Ang II intracellular molecular signaling that promotes insulin resistance in skeletal muscle cells through mitochondrial oxidative stress and Sirt3 dysfunction. It is conceivable that mechanism(s) at work in skeletal muscles may contribute to insulin resistance induced by Ang II in other tissues. The present study also highlights Sirt3 as a candidate therapeutic target for antioxidant and mitochondria-protective agents that counteract the deleterious effects of Ang II on insulin sensitivity and paves the way for testing novel treatments for insulin resistance, metabolic syndrome, and possibly diabetes, based on the pharmacological modulation of Sirt3.