Overexpression of SIRT1 in Rat Skeletal Muscle Does Not Alter Glucose Induced Insulin Resistance

SIRT1 is a NAD+-dependent deacetylase thought to regulate cellular metabolic pathways in response to alterations in nutrient flux. In the current study we investigated whether acute changes in SIRT1 expression affect markers of muscle mitochondrial content and also determined whether SIRT1 influenced muscle insulin resistance induced by acute glucose oversupply. In male Wistar rats either SIRT1 or a deacetylase inactive mutant form (H363Y) was electroprated into the tibialis cranialis (TC) muscle. The other leg was electroporated with an empty control vector. One week later, glucose was infused and hyperglycaemia was maintained at ~11mM. After 5 hours, 11mM glucose induced significant insulin resistance in skeletal muscle. Interestingly, overexpression of either SIRT1 or SIRT1 (H363Y) for 1 week did not change markers of mitochondrial content or function. SIRT1 or SIRT1 (H363Y) overexpression had no effect on the reduction in glucose uptake and glycogen synthesis in muscle in response to hyperglycemia. Therefore we conclude that acute increases in SIRT1 protein have little impact on mitochondrial content and that overexpressing SIRT1 does not prevent the development of insulin resistance during hyperglycaemia.


Introduction
Sirtuin 1 (SIRT1) is a NAD + -dependent deacetylase with a large range of target proteins that are important for apoptosis, the cell cycle, circadian rhythms, mitochondrial function, and metabolism [1]. SIRT1 is thought to be nutritionally regulated and be responsible for the beneficial effects of calorie restriction [2,3]. Levels of SIRT1 are reportedly decreased with highfat feeding and may therefore have a role in lipid-induced insulin resistance [4]. Consistent with these results, in vitro studies has shown that SIRT1 is down regulated under hyperglycaemic conditions in liver [5,6], endothelial [7,8], mesangial [9], corneal epithelial [10] and C2C12 muscle cells [11]. Rescuing the decrease in SIRT1 via pharmacological intervention (e.g. resveratrol) or protein overexpression can reverse the detrimental hyperglycaemic effect in these systems [8][9][10].
Using a model of hyperglycaemia (~11mM blood glucose) generated by a moderate intravenous glucose infusion into rats, we have shown previously that skeletal muscle insulin resistance consistently develops between 3 and 5h [12][13][14]. Interestingly, this insulin resistance developed prior to alterations in the insulin signaling pathway [12,13] but occurred in association with increased glycogen content and reduced AMPK activity [12,14]. In an attempt to further delineate the underlying mechanism(s), Saha et al [15] incubated EDL muscle strips in 5 or 25 mM glucose and found that the lactate to pyruvate ratio was increased, indicative of a decrease in the NAD + /NADH ratio. A tendency for SIRT1 protein levels to de decreased (20%, p = 0.10) was also found after incubation in this study [15]. Thus, combining the results from cell work and the muscle strip experiments, it is possible that SIRT1 may contribute to hyperglycemia-induced insulin resistance. Hence, the aim of the current study was to investigate whether SIRT1 overexpression would prevent the development of insulin resistance in skeletal muscle in vivo.

Cell culture
All cell culture reagents are from Life Technologies (Auckland, NZ) unless otherwise stated. Mouse skeletal muscle cell line C2C12 myoblasts were maintained in DMEM/F21 (1:1) with 10% FBS, 100 U/ml penicillin and 100 U/ml streptomycin (growth medium) at 37°C under a 5% CO 2 atmosphere. For transfection, the cells were grown on 6-well plates (Corning Inc., NY, USA) in 2 ml of growth medium. Once at 90% confluence, cells were transfected with either empty plasmid (2μg), a truncated but fully active form of SIRT1 [16], or a version of the SIRT1 construct that has a point mutation that renders it deacetylase inactive (SIRT1 (H363Y); kind gifts from Aimin Xu [16]) using the X-tremeGENE HP DNA transfection reagent (Roche, Mannheim, Germany) and opti-MEM reagent. Cells were collected 24h post transfection. Data presented are an average of 3 independent experiments.

Animals
All surgical and experimental procedures performed were approved by the Garvan Institute/St Vincent's Hospital Animal Ethics Committee and were in accordance with the National Health and Medical Research Council of Australia's guidelines on animal experimentation.
Adult male Wistar rats (Animal Resources Centre, Perth, Australia) were communally housed in temperature controlled (22 ± 0.5°C) 12 h light-dark cycle rooms. Rats were fed ad libitum a standard chow diet (Rat Maintenance Diet; Gordon Specialty Feeds, Sydney, Australia). Rats were acclimatized for 1 week prior to surgery.
In Vivo Electroporation (IVE) and Surgical Procedures. After the acclimatization period, rats were electroporated as previously described [17][18][19]. Briefly, under anaesthesia, control and test muscles were pretreated for 2 h with 90 units of hyaluronidase to break down components of the extracellular matrix to improve transfection efficiency [20]. Either SIRT1, or SIRT1 (H363Y), was injected into the test (right) tibialis cranialis (TC), and empty plasmid was injected into the control (left) TC via 6 x 50ul injections. Both legs underwent an electroporation protocol consisting of one 800 V/cm, 100 ms pulse followed by four 80 V/cm, 100 ms pulses at 1 Hz. Immediately after the IVE, while still under anaesthesia, dual cannulation of both jugular veins was performed as described previously [12,21].
Glucose Infusion. Seven days after surgery, rats (approximately 300g body weight) were randomly divided into treatment groups. After a basal blood sample was taken, a 50% (w/v) glucose infusion commenced. Rats were infused for either 0 or 5h using a peristaltic roller pump (101U/R; Watson-Marlow, Falmouth, UK). Blood samples were taken every 30 min and the glucose infusion rate was altered to maintain blood glucose concentration at~11 mM. Red blood cells from each sample were resuspended in heparinised saline and returned to the animal. 2-deoxy-D-[2,6-3 H]glucose and [U-14 C]glucose (PerkinElmer, Melbourne, Australia) were administered as an intravenous bolus in the last 30 min of the glucose infusion. Blood samples were taken 2, 5, 10, 15, 20 and 30 min after administration of the tracer bolus for estimation of tracer clearance and blood glucose. Animals were then euthanized and tissues were rapidly removed, freeze-clamped, and stored at -80°C for later analysis. TC muscle was powdered prior to any assay procedure to ensure homogeneity.

Analytic Methods
Blood and plasma glucose levels (YSI2300; Yellow Springs Instruments, Yellow Springs, OH, USA), and plasma insulin (Rat RIA, Millipore, Missouri, USA) were measured. Plasma and tissue levels of 3 H-and 14 C-labelled tracers were measured to calculate whole body glucose disposal rate (R d ), to estimate tissue glucose uptake (R g 0 ), and to measure glucose incorporation rate into glycogen. Assays and calculations for the glucose disappearance, glucose uptake into tissues and glucose incorporation into glycogen and glycogen content measures are as previously described [22]. Oxidation of palmitate and glutamate was assessed in muscle tissue homogenates as described previously [23]. Enzyme activities for citrate synthase (CS), β-hydroxyacyl CoA dehydrogenase (βHAD) and succinate dehydrogenase (SDH) were done as previously described [23,24]. Mitochondrial measurements were performed on animals in the basal state (without glucose infusion).

Statistics
Data are expressed as means ± SEM. Differences between groups were determined by paired students t-test, one-, or two-way ANOVA as appropriate (see figure legends). If the one-way ANOVA reached significance a Newman-Keuls multiple comparison post hoc test was conducted. If the two-way ANOVA reached significance a Bonferroni's post hoc test was conducted. All statistical analysis was performed using GraphPad Prism (Version 6 for Windows, GraphPad Software, San Diego, California, USA). The level of significance was p0.05.

In Vitro
To confirm the activity of the constructs, C2C12 myoblasts were successfully transfected with SIRT1, a deacetylase inactive mutant of SIRT1 (H363Y) or empty vector for 24h (Fig. 1A). The tumour suppressor protein p53 is a well-described target of SIRT1 [1]. There was a decrease in acetyl-p53 in the SIRT1 overexpressing myoblasts when compared to the empty vector and no change with the mutated version (Fig. 1B). Thus we confirmed the activity of the constructs and subsequently used them to increase SIRT1 protein in skeletal muscle of rats.

In Vivo
Basal. Electroporation of the SIRT1 and SIRT1 (H363Y) constructs significantly increased the levels of these proteins in the test leg (right) compared to the control leg (left; Fig. 2A, 2B).
To determine if overexpression of either of the SIRT1 constructs altered mitochondrial parameters we examined the content of the mitochondrial respiratory chain and porin, an abundant mitochondrial protein, often used as a marker of mitochondrial density [25,26] in control muscle and muscle overexpressing the SIRT1 constructs. There was no difference in these parameters in the left and right tibialis muscles ( Fig. 2C and D). As a measure of function, whole tissue homogenate oxidation rates of palmitate and glutamate were measured. Palmitate oxidation in tissue homogenates showed a small 11% increase in the SIRT1 overexpressing leg that was statistically significant using a paired t-test (p = 0.049). There was no change in the SIRT1 (H363Y) overexpressing leg (Fig. 2E). When glutamate was used no difference in substrate oxidation was observed in either the SIRT1 or SIRT1 (H363Y) overexpressing leg (Fig. 2F). Enzyme activities for CS, βHAD and SDH were all not different from the control leg for either SIRT1 or the deacetylase mutant form (Fig. 2G). This data suggests mitochondrial content and function were largely unaltered by the overexpression of either an active or inactive version of SIRT1 for 1 week.   Glucose Infusion. Blood glucose levels increased significantly in glucose-infused animals and remained stable over the 5h infusion period (Fig. 3A). Plasma insulin was also elevated and remained stable during glucose infusion (Fig. 3B). The amount of glucose infused to maintain hyperglycemia decreased between 3h and 5h of infusion, indicating the existence of wholebody insulin resistance at 5h (Fig. 3C).
After 5h glucose infusion, tibialis muscle glucose uptake was assessed as previously described [12][13][14]. Overexpression of SIRT1 or SIRT1 (H363Y) for 7 days had no effect on glucose uptake into tibialis muscle (Fig. 4A). There were also no differences between the test and control legs in glycogen synthesis or content after 5h infusion, irrespective of which construct was overexpressed ( Fig. 4B and C).
Immunoblotting for phosphorylation status of insulin receptor (IR) or Akt showed that neither construct altered basal phosphorylation status of these proteins, nor the increase seen in response to the infusion (Fig. 5A and B). In the current study, there was also no difference in the basal phosphorylation state of ACC or the decrease in phosphorylation of ACC in response  to hyperglycaemia in the SIRT1, or SIRT1 (H363Y), over expressing leg (Fig. 5C). This indicates that AMPK activity was unlikely to be affected by overexpression of SIRT1.

Discussion
The current study provides evidence that increasing SIRT1 expression has no obvious impact on the development of glucose-induced insulin resistance in skeletal muscle. When SIRT1 or the deacetylase inactive version of SIRT1 (H363Y) were over expressed in muscle, there was no alteration of the insulin resistance that occurs with 5h of hyperglycaemia [12][13][14]. In previous in vitro studies, an increase in SIRT1 protein or activity was associated with a rescuing of the effects of incubation with high glucose (e.g. oxidative damage or wound healing [9,10]). One possibility for the differences between the previous in vitro and current in vivo studies is that in vivo activity of SIRT1 is tightly regulated by intracellular NAD + content. Under such conditions it is possible that a decreased activity of SIRT1 cannot be overcome by simply overexpressing the protein. Other possibilities for the apparently different effects of SIRT1 overexpression are the origin of the cells used for in vitro studies (mesangial cells [9] or endothelial cells [8,10]), in vivo vs in vitro conditions, and/or duration of glucose exposure. Most studies, especially those in cells, use at least 24h of the hyperglycaemic insult, whereas our study examined 5h glucose infusion.
It has been reported previously that SIRT1 can modulate the insulin signaling pathway via repression of PTP1B [4], Rictor of the mTORC2 complex [27], or via regulation of the p85 [28], p55α and p50α [29] subunits of PI3K. In the current study we did not find any differences in the phosphorylation status of the insulin receptor, a target of PTP1B, or Akt, which is downstream of both mTORC2 and PI3K, either in the basal state or in response to hyperglycaemia. This indicates that one week of overexpression of SIRT1 protein is insufficient to modulate these indicators of insulin signaling pathway activity although this does not preclude that longer term increases in SIRT1 protein may have some effects.
Similarly, SIRT1 has been reported to regulate AMPK signaling through deacetylation and activation of its upstream kinase LKB1 [30]. Conversely, AMPK is thought to regulate SIRT1 activity by modulating NAD + levels [31]. However phosphorylation of ACC, a well described target of AMPK, was not changed by overexpression of SIRT1 in either the basal state or in response to hyperglycaemia, suggesting unaltered AMPK signaling in this model. This lack of effect of SIRT1 on AMPK and ACC is consistent with other studies reporting no difference in the phosphorylation state or activity of AMPK and ACC when SIRT1 is specifically knocked out [29,32], or overexpressed [33] in muscle.
Consistent with previous in vitro studies associating SIRT1 with fatty acid oxidation [34][35][36], we found a small increase in palmitate oxidation using homogenates prepared from the SIRT1 overexpressing (test) leg. The mechanism mediating this increase remains unclear although mitochondrial content and metabolism in the basal state were unaffected by the increase in SIRT1 protein.
Although there is evidence for a role of SIRT1 in mitochondrial biogenesis, we show that there was also no change in markers of mitochondrial content, or function, in muscle after overexpression of SIRT1 for one week. Recently however, there has been some controversy about the role SIRT1 plays in mitochondrial biogenesis in muscle, especially in vivo. Supporting its role, a comprehensive study in C2C12 muscle cells showed that SIRT1, through its ability to deacetylate and activate PGC-1α, was intimately linked to mitochondrial biogenesis and function [35]. In vivo, studies in conditional whole body knockout [37], muscle specific knockout [38] as well as whole body over expression [37] of SIRT1 show similar links. Other studies by Philp et al [32] reported that muscle-specific knockout of SIRT1 does not alter the abundance or activities of complexes within the respiratory chain, or mitochondrial content. Interestingly though, when SIRT1 was overexpressed using a similar electroporation protocol to ours, there was a decrease in components of the respiratory chain [39,40] while a dominantnegative version of SIRT1 (H355Y) of SIRT1 increased these components [40]. More recently, transgenic overexpression of SIRT1 has reportedly no alteration in respiration rates of isolated mitochondria, with only small increases in some components of the respiratory chain (complex 2 and 4, cyt c) and no protection against the effects of a high fat diet [41]. Thus, the role of SIRT1 in regulating muscle mitochondrial function under normal conditions as well as in response to exercise or resveratrol treatment will remain controversial until the results of more studies using similar paradigms become available [42][43][44].
In a cultured cell system we showed that overexpression of our SIRT1 construct led to a reduced acetylation of p53 using an antibody directed against acetylated p53 (K379). This demonstrated that the construct produced a functional SIRT1 enzyme capable of altering protein acetylation. In control and SIRT1 overexpressing TC muscle the same antibody did not detect any p53 acetylation and therefore there is no direct evidence that the SIRT1 protein produced in TC muscle was capable of altering protein acetylation. However, we have used the electroporation protocol previously to overexpress many functional proteins [17][18][19] and see no reason to suspect that the SIRT1 enzyme would not be active in TC muscle if substrate availability and other potential regulatory factors were favourable. It is possible that tight regulation of SIRT1 activity via substrate (NAD + ) availability occurs in vivo, explaining the lack of effect of simply increasing the amount of SIRT1 protein.
In conclusion, our results suggest that although a decrease in SIRT1 may be involved in the development of insulin resistance in skeletal muscle in response to hyperglycaemia, overexpressing SIRT1 does not alleviate this insulin resistance. This data also shows that increasing SIRT1 protein has little effect on muscle mitochondrial biogenesis in vivo. The acetylation/ sirtuin axis is clearly complex and changes in SIRT1 protein may not equate to linear changes in enzyme activity, due to requirement for NAD + as a co-factor. Therefore strategies aimed at raising NAD + levels [45][46][47] might be more useful in examining the role of SIRT1 in the acute regulation of muscle metabolism in response to a hyperglycaemic insult.