Increased glucose production and reduced hepatic glycogen storage contribute to metabolic abnormalities in diabetes. Irisin, a newly identified myokine, induces the browning of white adipose tissue, but its effects on gluconeogenesis and glycogenesis are unknown. In the present study, we investigated the effects and underlying mechanisms of irisin on gluconeogenesis and glycogenesis in hepatocytes with insulin resistance, and its therapeutic role in type 2 diabetic mice. Insulin resistance was induced by glucosamine (GlcN) or palmitate in human hepatocellular carcinoma (HepG2) cells and mouse primary hepatocytes. Type 2 diabetes was induced by streptozotocin/high-fat diet (STZ/HFD) in mice. In HepG2 cells, irisin ameliorated the GlcN-induced increases in glucose production, phosphoenolpyruvate carboxykinase (PEPCK) and glucose-6-phosphatase (G6Pase) expression, and glycogen synthase (GS) phosphorylation; it prevented GlcN-induced decreases in glycogen content and the phosphoinositide 3-kinase (PI3K) p110α subunit level, and the phosphorylation of Akt/protein kinase B, forkhead box transcription factor O1 (FOXO1) and glycogen synthase kinase-3 (GSK3). These effects of irisin were abolished by the inhibition of PI3K or Akt. The effects of irisin were confirmed in mouse primary hepatocytes with GlcN-induced insulin resistance and in human HepG2 cells with palmitate-induced insulin resistance. In diabetic mice, persistent subcutaneous perfusion of irisin improved the insulin sensitivity, reduced fasting blood glucose, increased GSK3 and Akt phosphorylation, glycogen content and irisin level, and suppressed GS phosphorylation and PEPCK and G6Pase expression in the liver. Irisin improves glucose homoeostasis by reducing gluconeogenesis via PI3K/Akt/FOXO1-mediated PEPCK and G6Pase down-regulation and increasing glycogenesis via PI3K/Akt/GSK3-mediated GS activation. Irisin may be regarded as a novel therapeutic strategy for insulin resistance and type 2 diabetes.
Increased glucose production and reduced hepatic glycogen storage contribute to metabolic abnormalities in diabetes. The newly identified myokine irisin induces the browning of WAT and increases energy expenditure. However, the effects of irisin on hepatic glycogenesis and gluconeogenesis are unknown.
In the present study, irisin reduces gluconeogenesis via the PI3K/Akt/FOXO1-mediated down-regulation of PEPCK and G6Pase, and increases glycogenesis via the PI3K/Akt/GSK3-mediated GS activation in the human HepG2 cell line and mouse primary hepatocytes with insulin resistance, as well as in type 2 diabetic mice. Moreover, long-term subcutaneous administration of irisin attenuates hyperglycaemia and insulin resistance in mice with type 2 diabetes.
This study provides evidence that irisin inhibits hepatic gluconeogenesis, increases glycogen synthesis and improves insulin resistance. Irisin may be regarded as a novel therapeutic strategy for insulin resistance and type 2 diabetes.
Type 2 diabetes is characterized by hyperglycaemia and insulin resistance in target tissues. The liver plays a central role in the maintenance of blood glucose by balancing new synthesis (gluconeogenesis), glucose uptake and storage via glycogen synthesis (glycogenesis) and glucose release via the breakdown of glycogen (glycogenolysis) . Glycogen synthesis is mainly regulated by glycogen synthase kinase-3 (GSK3) and glycogen synthase (GS) . Gluconeogenesis is primarily modulated by phosphoenolpyruvate carboxykinase (PEPCK) and glucose-6-phosphatase (G6Pase) , which are up-regulated in response to hormones during fasting, and robustly down-regulated by insulin and glucose [1,4]. Gluconeogenesis is strongly stimulated during fasting and is aberrantly activated in diabetes . Excessive hepatic glucose production is a major contributor to both fasting hyperglycaemia and exaggerated postprandial hyperglycaemia in types 1 and 2 diabetes . Insulin resistance in the liver is an important factor causing fasting hyperglycaemia and type 2 diabetes .
Irisin is a cleaved and secreted fragment of fibronectin type III domain-containing protein 5 (FNDC5), which induces uncoupling protein-1 (UCP1) expression and white adipose tissue (WAT) browning. This newly identified myokine enhances energy expenditure and may be involved in the beneficial effects of exercise that have been noted independent of weight loss [7,8]. Adenoviral vectors that express full-length FNDC5 in mice improve glucose tolerance and reduce insulin resistance . Circulating irisin was positively correlated with age, body mass index (BMI), total cholesterol and triacylglycerols, and fasting blood glucose in non-diabetic individuals, and the increased circulating irisin may be an adaptive response to compensate for the decreasing insulin sensitivity and disturbances in metabolism associated with obesity . However, serum irisin levels were similar in pregnant women from the non-obese, obese and gestational diabetes groups, as well as in a separate cohort of non-pregnant, lean and obese women. Analysis of the pregnant women revealed that there was a significant inverse correlation between BMI and serum irisin .
The controversy over the serum irisin levels in various studies may be related to the gender difference of irisin levels in various studies . Although some studies show conflicting results, most show that serum irisin levels are lower in patients with type 2 diabetes [12–15]. It is noted that strong irisin immunoreactivity was found in the liver . Serum irisin concentrations had an inverse association with the triacylglycerol content in the liver and liver enzymes in obese adults . Previous studies have shown that insulin-induced suppression of glycogenolysis and gluconeogenesis was impaired in obesity and type 2 diabetes . Defects in glycogenolysis and gluconeogenesis have been shown to be involved in hepatic insulin resistance in humans with type 2 diabetes . However, the effects of irisin on hepatic glycogenesis and gluconeogenesis are unknown.
In the present study, the effects of irisin on glycogenesis and gluconeogenesis and the related signal pathway were investigated in human hepatocellular carcinoma (HepG2) cells and mouse primary hepatocytes with insulin resistance; the therapeutic effects of persistent administration of irisin on glucose metabolic disturbance and insulin resistance were determined in mice with type 2 diabetes induced by streptozotocin (STZ)/high-fat diet (HFD).
MATERIALS AND METHODS
HepG2 cell culture and insulin resistance models
HepG2 cells were obtained from the American Type Culture Collection, and maintained in Dulbecco's modified Eagle's medium (DMEM) with 25 mM glucose, 10% FBS, penicillin (100 units/ml) and streptomycin (100 μg/ml) at 37°C in a humidified atmosphere of 95% air and 5% CO2. To induce insulin resistance, HepG2 cells were incubated with 18 mM glucosamine (GlcN) for 18 h or 0.25 mM palmitate for 24 h in serum-free medium [20,21], followed by incubation in medium with 20 nM irisin or vehicle (PBS) for 30 min for measurement of phosphorylated protein, or 24 h for other measurements in the presence of GlcN or palmitate. The phosphoinositide 3-kinase (PI3K) inhibitor LY294002 (10 μM) and the Akt/protein kinase B inhibitor MK2206 (1 μM) were added into the medium 24 h before administration of irisin or vehicle for assessing the impact of these inhibitors. Cell viability was detected using the Cell Counting Kit-8 (CCK8) assay (Dojindo Molecular Technologies), according to the manufacturer's instructions.
Isolation of primary hepatocytes
Male C57BL/6J mice were anaesthetized by intraperitoneal injection of pentobarbital (50 mg/kg). Perfusion was made via the portal vein with 50 ml of Hepes buffer containing collagenase II (0.66 mg/ml) at 37°C. The liver was aseptically excised. Cells were dispersed and filtrated via a 70-μm cell strainer into a centrifuge tube. After centrifugation, cells were washed and cell viability was examined using Trypan Blue dye before >95% were used for subsequent experiments. The hepatocytes were maintained in low-glucose DMEM containing 10% FBS, with penicillin and streptomycin for 24 h before inducing insulin resistance with GlcN.
The medium in six-well plates was replaced with 2 ml of glucose production buffer consisting of glucose-free DMEM without Phenol Red, supplemented with 20 mM sodium lactate and 2 mM sodium pyruvate. After 3 h of incubation, half of the glucose production buffer was collected and the glucose concentration was measured using a glucose oxidase–peroxidase assay kit (Jiancheng Bioengineering Institute). The values were normalized to the total protein content, determined from the whole-cell extracts .
Glycogen levels were tested using a Glycogen Assay Kit (BioVision) following the manufacturer's instructions. Briefly, cells or tissues were homogenized with 100 μl of ice-cold glycogen development buffer for 10 min. Homogenates were spun at 15000 g for 5 min and supernatants were assayed for glycogen content. Glycogen concentration was normalized by cell numbers (109 in each plate) or protein concentration of tissues .
Quantitative real-time PCR
Total RNA was isolated from cells or tissues using TRIzol Reagent (Invitrogen). Real-time PCR was performed using a StepOnePlus Real-Time PCR System (Applied Biosystems). All gene expression levels were normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) levels. The sequences of primers are listed in Supplementary Table S1.
Protein extracts were electrophoresed, blotted and then incubated with antibodies against Akt, phospho-Akt (Ser473), PI3K p110α, PI3K p110β, GS, phospho-GS (Ser641), forkhead box transcription factor O1 (FOXO1), phospho-FOXO1 (Ser256), GSK3, phospho-GSK3 (Ser9), GAPDH (Cell Signaling Technology), PEPCK (Abcam) and G6Pase (Santa Cruz Biotechnology) with appropriate secondary horseradish peroxidase-conjugated antibodies, and then developed.
G6Pase and PEPCK activity analysis
G6Pase activity and PEPCK activity were measured as previously described . Briefly, G6Pase activity was assayed by quantifying the released phosphate. Phosphoenolpyruvate was carboxylated by PEPCK to form oxaloacetate, which was then converted into malate using malic dehydrogenase. This conversion was monitored by a decrease in NADH absorbance at 340 nm.
Mice model of type 2 diabetes
Male C57BL/6J mice aged 6 weeks were used for inducing type 2 diabetes. Experiments were approved by the Experimental Animal Care and Use Committee of Nanjing Medical University and conducted in accordance with the Guide for the Care and Use of Laboratory Animals (NIH publication, 8th edition, 2011). The mice were caged in an environment under controlled temperature and humidity, with free access to water and diet under a 12-h light/12-h dark cycle. A combination of rats or mice treated with low-dose streptozotocin (STZ) or a high-fat diet (HFD) is identified as an ideal animal model for type 2 diabetes, because it simulates the natural disease progression and metabolic characteristics typical of type 2 diabetes [25,26]. This model is appropriate for testing anti-diabetic agents in the treatment of type 2 diabetes . After a week of adaptation in the living environment, the mice were randomly divided into three groups (n=7 for each group). One group of mice (Ctrl) received an intraperitoneal injection of vehicle and were fed with a normal diet (14.7 kJ/g, 13% of energy as fat) throughout the experiment. Another two groups of mice were combined as one group (STZ/HFD), to induce type 2 diabetes mellitus. These mice were subjected to 4-h fasting followed by intraperitoneal injection of low-dose STZ (120 mg/kg body weight in 10 mmol/l citrate buffer, pH 4.0). After 3 weeks, the mice were fed with a HFD (21.8 kJ/g, 60% of energy as fat–D12492, Research Diets) instead of the previously used normal diet. The mice were re-divided into two groups, 8 weeks after the injection of STZ, and received, respectively, subcutaneous perfusion of either saline or irisin via micro-osmotic pumps for 2 weeks, and were maintained on HFD feeding (see Supplementary Figure S1).
Subcutaneous perfusion of irisin
Mice were anaesthetized with an intraperitoneal injection of pentobarbital (50 mg/kg) to implant the pump. A micro-osmotic pump (model 1002, Alzet) was implanted into the mid-scapular region of each mouse under aseptic manipulation. The pump delivered saline or irisin (0.24 nmol/μl, 1.44 nmol/day) for 14 days.
Measurement of insulin and irisin levels
Serum insulin level was determined using a mouse insulin ELISA kit (ALPCO). Serum and liver irisin levels were measured with an irisin human, rat, mouse and canine enzyme immunoassay (EIA) kit (EK-067-16, Phoenix Pharmaceuticals). For the anti-irisin antibody used in the present study, the minimum detectable concentration of irisin is 6.8 ng/ml. The cross-reactivity is 100% with irisin or irisin (42-112), and 9% with FNDC. No cross-reactivity was found with FNDC5 (165-212), FNDC5 (162-209) or irisin (42-95). The recovery was 104%, 82% and 88.7% for spiked irisin at 5, 10 and 20 ng/ml, respectively. Intra- and inter-assay variations are <10% and <15%, respectively. The accuracy and validity have been checked . The inter- and intra-observer variabilities of measurements with the kits were 6.2% and 5.9%, respectively . We conducted a validation of irisin (spiking and recovery) for the ELISA kit, and found that the recovery was 101.8%, 98.0% and 88.2% for spiked irisin at 5, 10 and 20 ng/ml, respectively.
Insulin tolerance test and glucose tolerance test
For the insulin tolerance test (ITT) and the glucose tolerance test (GTT), mice were starved for 6 h and overnight, respectively, before an intraperitoneal injection of insulin (0.75 units/kg of body weight) or glucose (2.0 g/kg of body weight) . Glucose levels in tail-vein blood samples were measured using a blood glucometer (OneTouch) at 15, 30, 60 and 120 min after the injection.
Detection of glycogen deposits in liver
Periodic acid–Schiff (PAS) staining was used to visualize the glycogen deposits in livers. Liver tissue was fixed with 4% paraformaldehyde and then mounted in paraffin and sectioned. Sections were oxidized in periodic acid for 20 min at room temperature and rinsed three times in deionized water. After the treatment with 0.5% sodium bisulfite in 0.05 M HCl for 10 min, sections were counterstained with haematoxylin and eosin for 1 min, rinsed in deionized water and observed for the red-staining parts as the presence of glycogen using a light microscope.
STZ, collagenase II and palmitate were purchased from Sigma. Glucose and glucosamine were obtained from Beyotime Institute of Biotechnology (Bio-Equip). LY294002 and MK2206 were bought from Selleck Chemicals. Irisin was bought from ChinaPeptides, which was derived from Escherichia coli, and was purified and identified. The quality and quantity of purified irisin were analysed using SDS/PAGE. Irisin was treated with a Pierce high-capacity endotoxin-removal resin (Thermo Scientific) to remove any endotoxin.
Values were expressed as the means±S.E.M. One-way or two-way ANOVA followed by post-hoc Bonferroni's test was used for the multiple comparisons. Student's t test was used for comparisons between the two groups. A value of P<0.05 was considered statistically significant.
Gluconeogenesis and glucose production in hepatocytes
GlcN or palmitate is used to induce insulin resistance in vitro in HepG2 cells and primary hepatocytes [20,21]. PEPCK and G6Pase are two key hepatic gluconeogenesis enzymes . GlcN treatment increased PEPCK and G6Pase mRNA levels (P=0.011 and 0.012), protein expression (P=0.004 and 0.049) and activity (P=0.024 and 0.037), as well as glucose production (P=0.023) in HepG2 cells. These findings were not seen when the cells were incubated in the presence of irisin (Figure 1). Similar effects of irisin on PEPCK and G6Pase mRNA expression (see Supplementary Figure S2) and glucose production (see Supplementary Figure S3) were found in GlcN-treated mouse primary hepatocytes. Furthermore, effects of irisin on glucose production were confirmed in palmitate-treated HepG2 cells (see Supplementary Figure S4). These results indicate that irisin rectifies the enhanced gluconeogenesis and glucose production in hepatocytes with insulin resistance.
Effects of irisin on the key enzymes of gluconeogenesis (PEPCK and G6Pase) and glucose production in GlcN-treated human HepG2 cells
Glycogen synthesis and glycogen content in hepatocytes
GSK3 inhibits glycogen synthesis via phosphorylation of GS, a key enzyme that catalyses the last step in glycogen synthesis . Insulin is known to increase GSK3 phosphorylation and thereby activates GS, causing the conversion of glucose to glycogen . GlcN treatment reduced GSK3 phosphorylation (P=0.042) and glycogen content (P=0.020), but increased GS phosphorylation (P=0.003); these were ameliorated by irisin in human HepG2 cells (P=0.032, 0.039 and 0.047) (Figure 2). Similar effects of irisin on glycogen synthesis (see Supplementary Figure S5) and glycogen content (see Supplementary Figure S6) were observed in GlcN-treated mouse primary hepatocytes. Furthermore, effects of irisin on glycogen content were confirmed in palmitate-treated HepG2 cells (see Supplementary Figure S4). The findings provide ample evidence that irisin prevents reduced glycogen synthesis and glycogen content in hepatocytes with insulin resistance.
Effects of irisin on the key enzymes of glycogenesis (GSK3 and GS) and glycogen level in GlcN-treated human HepG2 cells
PI3K/Akt/FOXO1 pathway in hepatocytes
The enzyme activity of PI3K is mediated by a group of catalytic and regulatory subunits . FOXO1 promotes the expression of PEPCK and G6Pase, and its phosphorylation causes its inactivation in hepatocytes . Insulin suppresses FOXO1 activity through activation of the PI3K/Akt signalling pathway . We found that both p110α and p110β subunits of PI3K were down-regulated in GlcN-treated HepG2 cells (P=0.017 and 0.009), but irisin prevented expression of only the reduced p100α subunit (P=0.013), rather than that of the p100β subunit (P=0.153) (Figures 3A and 3B). Moreover, the phosphorylation of Akt and FOXO1 was reduced in GlcN-treated HepG2 cells (P=0.040 and 0.031), which were eliminated by irisin (P=0.003 and 0.045) (Figures 3C and 3D). These results indicate that irisin causes FOXO1 phosphorylation, thereby inhibiting FOXO1 activity via activation of the PI3K/Akt signalling pathway. It is well known that insulin inhibits gluconeogenesis and increases glycogen synthesis via the PI3K/Akt pathway [36,37]. We therefore found that irisin showed similar phosphorylation effects on Akt and FOXO1 to insulin in control HepG2 cells (Figures 3F and 3G), suggesting that irisin and insulin share the PI3K/Akt pathway.
Effects of irisin on the PI3K p110
α subunit and p110 β subunit expression, and the phosphorylation of Akt and FOXO1
Effects of PI3K and Akt inhibitors in hepatocytes
The PI3K inhibitor LY294002 prevented the role of irisin in reducing PEPCK and G6Pase mRNA levels (P=0.026 and 0.067), activity (P=0.041 and 0.032) and protein expression (P=0.012 and 0.001), as well as glucose production (P=0.002) in GlcN-treated HepG2 cells. The Akt inhibitor MK2206 showed a similar role in preventing these effects of irisin in GlcN-treated HepG2 cells (Figures 4A and 4D, and see Supplementary Figures S7 and S8). Inhibition of PI3K or Akt abolished the role of irisin in increasing the phosphorylation of FOXO1 (P=0.0001 and 0.0001) in GlcN-treated HepG2 cells (Figure 4F), indicating that irisin prevents increased gluconeogenesis and glucose production in hepatocytes with insulin resistance via PI3K/Akt-mediated FOXO1 inactivation, and subsequent inhibition of PEPCK and G6Pase mRNA expression. On the other hand, LY294002 or MK2206 abolished the role of irisin in promoting GSK3 phosphorylation (P=0.022 and 0.010) (Figure 4B), reducing GS phosphorylation (P=0.002 and 0.004) (Figure 4C) and increasing the glycogen content (P=0.004 and 0.094) (Figure 4E) in GlcN-treated HepG2 cells; this indicates that irisin normalizes reduced glycogen synthesis and glycogen content in hepatocytes with insulin resistance via PI3K/Akt-mediated GSK3 inactivation and subsequent GS activation. The findings were further confirmed in GlcN-treated mouse primary hepatocytes in which LY294002 or MK2206 abolished the role of irisin in reducing glucose production (see Supplementary Figure S9) and increasing glycogen content (see Supplementary Figure S10).
PI3K inhibitor LY294002 or Akt inhibitor MK2206 prevented the effects of irisin on PEPCK, G6Pase, GSK3, glucose production, glycogen and FOXO1 in GlcN-treated human HepG2 cells
Irisin levels, body weight and food intake in mice
The effects of the subcutaneous perfusion of irisin with micro-osmotic pumps on glucose metabolism and insulin resistance were identified in mice with type 2 diabetes induced by STZ/HFD. The serum and liver irisin levels were lowered in STZ/HFD mice (P=0.035 and 0.047), and irisin perfusion increased serum and liver irisin levels in STZ/HFD mice (P=0.004 and 0.020) (Figures 5A and 5B), confirming the effectiveness of irisin perfusion using osmotic micro-pumps. However, subcutaneous perfusion of irisin had no significant effect on body weight, food intake and systolic blood pressure (Figures 5C–5F).
Effects of irisin on irisin levels, body weight, food intake and blood pressure in Ctrl and STZ/HFD mice
Glucose metabolism in mice
STZ/HFD mice showed much higher fasting blood glucose levels than control mice (P=0.020). STZ/HFD mice did not show a significant change in serum insulin levels (P=0.166), and irisin perfusion had no significant effect on serum insulin levels in STZ/HFD mice (P=0.084) (Figure 6A). Irisin perfusion reduced the increased fasting blood glucose level (P=0.032) in STZ/HFD mice (Figure 6B). Liver glycogen content was reduced in STZ/HFD mice (P=0.0001), which was attenuated by the irisin treatment (P=0.007) (Figure 6C). Similarly, liver sections with PAS staining showed that the reduced glycogen in STZ/HFD mice was rectified by irisin perfusion (Figure 6D). GTT and ITT were used for evaluating glucose tolerance and insulin resistance. In the GTT, STZ/HFD mice manifested significantly elevated glucose excursions after a glucose challenge compared with control mice. The glucose excursion was reduced in irisin-treated STZ/HFD mice (Figure 6E). In the ITT, the efficiency of insulin was quantified by its ability to reduce blood glucose levels. Insulin was less effective in STZ/HFD mice than in control mice, suggesting impaired insulin sensitivity in STZ/HFD mice. Irisin improved insulin sensitivity in the STZ/HFD mice (Figure 6F). These results indicate that persistent administration of irisin reduces blood glucose levels, increases liver glycogen synthesis and attenuates insulin resistance in mice with type 2 diabetes.
Effects of irisin on glucose metabolism and insulin resistance in Ctrl and STZ/HFD mice
Hepatic gluconeogenesis and glycogen synthesis, and Akt phosphorylation in mice
Irisin perfusion attenuated the up-regulation of PEPCK and G6Pase mRNA levels (P=0.045 and 0.042), protein expression (P=0.049 and 0.002) and activity (P=0.028 and 0.013) in the livers of STZ/HFD mice (Figures 7A–7C). GSK3 phosphorylation was reduced, whereas GS phosphorylation was increased in STZ/HFD mice, which were normalized by irisin perfusion (P=0.002 and 0.023) (Figures 7D and 7E). Furthermore, Akt phosphorylation was inhibited in STZ/HFD mice, but was recovered by irisin perfusion (P=0.005) (Figure 7F). The in vivo experiments confirmed our in vitro findings that irisin prevents enhanced gluconeogenesis and decreased glycogen synthesis via the Akt-mediated regulation of key enzymes of gluconeogenesis and glycogen synthesis in diabetes.
Effects of irisin on PEPCK, G6Pase, GSK3, GS and Akt in the livers of Ctrl and STZ/HFD mice
The liver is primarily responsible for the maintenance of blood glucose levels by its ability to produce glucose from gluconeogenesis or glycogen breakdown and to store glucose as glycogen . Increased endogenous glucose production and reduced hepatic glycogen storage contribute to the metabolic abnormalities in diabetes . The primary novel findings of the present study are that irisin reduces gluconeogenesis via PI3K/Akt/FOXO1-mediated PEPCK and G6Pase down-regulation and increases glycogenesis via PI3K/Akt/GSK3-mediated GS activation in hepatocytes with insulin resistance. Another important finding is that persistent administration of irisin effectively reduces fasting blood glucose levels, hepatic gluconeogenesis and glucose production, increases hepatic glycogen synthesis and storage, and improves insulin resistance in mice with STZ/HFD-induced type 2 diabetes.
It is well known that increased endogenous glucose production, primarily of hepatic origin, is a major determinant of fasting hyperglycaemia in type 2 diabetes . Gluconeogenesis is enhanced in diabetes, and hepatic insulin resistance in diabetes is linked to a reduced ability of insulin to inhibit gluconeogenesis and glucose production in the liver . Reduction of hepatic glucose production can be considered a therapeutic target in diabetes [2,18]. Type 2 diabetes is strongly associated with a decrease in insulin-stimulated GS activity and glycogen synthesis . The present study provides evidence that irisin inhibited hepatic gluconeogenesis and endogenous glucose production, and increased hepatic glycogen synthesis and storage, not only in human HepG2 cells and mice primary hepatocytes with insulin resistance, but also in mice with type 2 diabetes. The decreased gluconeogenesis and increased glycogenesis at least partially contribute to the irisin-induced reduction in fasting blood glucose levels and the improvement in glucose homoeostasis and insulin resistance in type 2 diabetes. These results suggest that irisin may be an effective therapeutic strategy for type 2 diabetes.
Hepatic glucose production is mainly regulated by PEPCK and G6Pase, which are important target genes of FOXO1 and catalyse the rate-limiting steps in gluconeogenesis  (Figure 8). Activation of FOXO1 in liver induces gluconeogenesis via increasing PEPCK and G6Pase expression, and insulin suppresses hepatic gluconeogenesis via activation of the PI3K/Akt pathway and subsequent FOXO1 phosphorylation and inactivation [34,40]. In the present study, we found that irisin reduces gluconeogenesis via PI3K/Akt/FOXO1-mediated PEPCK and G6Pase down-regulation, which is similar to the signal pathway of insulin in the suppression of gluconeogenesis.
Schematic diagram showing the downstream signal mechanism of irisin in modulating gluconeogenesis and glycogen synthesis in hepatocytes
GS is known to catalyse the rate-limiting step for insulin-mediated glycogen synthesis, and GSK3 inhibits glycogen synthesis by suppressing GS via inhibitory phosphorylation . GSK3 has been implicated as mediating the development of insulin resistance, mainly by inhibition of glycogen synthesis . Insulin activates the PI3K/Akt signalling cascade and subsequent GSK3 phosphorylation and inhibition. This GSK3 inhibition leads to GS activation and thereby glycogen synthesis . GSK3 inhibitors may have a therapeutic potential in the treatment of diabetes and insulin resistance . In the present study, we found that irisin increases liver glycogen synthesis via PI3K/Akt/GSK3-mediated GS activation in insulin resistance and type 2 diabetes, which is similar to the signal pathway of insulin in the promotion of glycogen synthesis.
It is very interesting that irisin and insulin share a similar downstream signal pathway in reducing gluconeogenesis and increasing glycogen synthesis in insulin resistance and type 2 diabetes. The present in vitro and in vivo studies indicate that irisin is effective in preventing excess gluconeogenesis and reduced glycogen synthesis in insulin resistance and type 2 diabetes. More importantly, persistent administration of irisin reduces fasting blood glucose levels and improves insulin resistance in mice with STZ/HFD-induced diabetes. The greatest strength is that the irisin is effective for treating diabetes with insulin resistance. Although insulin and irisin may share the same pathway, the side effect of weight gain may be avoided if irisin is used as a therapeutic agent in type 2 diabetes.
Insulin regulates gluconeogenesis and glycogen synthesis by insulin receptor substrate phosphorylation mediated by the insulin receptor, thereby activating the PI3K/Akt signal pathway . However, the molecular mechanisms of irisin in inducing PI3K/Akt activation remain unknown. A limitation of the present study is that the irisin receptors and the interaction of irisin and insulin are not determined; this is worth further investigation. It is noted that irisin had no significant effect on serum insulin levels, indicating that the effect of irisin is not caused by a change in insulin levels. Moreover, the effects of irisin on glycogenolysis, glucose uptake of cells and lipid metabolism were not investigated in the present study. These effects may contribute to the beneficial effects of irisin in treating diabetes.
It has been found that circulating irisin correlated positively with BMI in non-diabetic individuals [43,44]. The circulating irisin level was higher in morbidly obese individuals than normal weight and anorexic patients . Weight loss induced by bariatric surgery reduced circulating irisin levels . The increased circulating irisin could be an adaptive response to compensate for the decreasing insulin sensitivity and disturbances in metabolism associated with obesity . It is noteworthy that circulating irisin levels were reduced in patients with type 2 diabetes [12–15]. In the present study, reduced serum and liver irisin levels were found in mice with STZ/HFD-induced type 2 diabetes. Administration of irisin improved the glucose metabolism in type 2 diabetic mice and in hepatocytes with insulin resistance. These results indicate that exogenous irisin is effective in ameliorating glucose metabolic derangements in diabetic mice with insulin resistance.
It has been shown that most previously published assays based on commercial ELISAs were reporting unknown cross-reacting proteins, and the ELISAs for irisin do not measure irisin, calling into question the role of irisin. It was found that commercial anti-irisin antibodies can measure recombinant non-glycosylated irisin, but not recombinant glycosylated irisin . So far, however, all previous studies showed that irisin is a bioactive molecule, and no study showed that irisin plays its role via its glycosylation. It is possible that glycosylation blocks the binding site of antibody to the glycosylated irisin. Albrecht et al.  made a comparison of ELISA data with Western blot analyses in four of five commercial antibodies except EK-067-16, and thus their conclusion is not completely applicable to EK-067-16. Most importantly, the cross-reactivity of the anti-irisin antibody (EK-067-16) used in the present study is known, and its accuracy and validity have been checked by the company and independent researchers [28,29]. The spiking-and-recovery test in our laboratory supports the validation of this commercial irisin ELISA kit.
In conclusion, irisin decreases gluconeogenesis by the PI3K/Akt/FOXO1-mediated down-regulation of PEPCK and G6Pase, and increases glycogenesis by the PI3K/Akt/GSK3-mediated GS activation in the human HepG2 cell line, mouse primary hepatocytes and mouse models of type 2 diabetes. Importantly, long-term subcutaneous administration of irisin attenuates hyperglycaemia and insulin resistance in mice with type 2 diabetes. The effects of irisin on gluconeogenesis and glycogen synthesis at least partially contribute to reduced glucose production, increased glycogen accumulation and improved glucose homoeostasis in type 2 diabetes. Irisin may be taken as an effective therapeutic strategy for type 2 diabetes.
Tong-Yan Liu, Chang-Xiang Shi, Run Gao, Hai-Jian Sun, Xiao-Qing Xiong and Lei Ding performed the experimental study. Tong-Yan Liu, Qi Chen, Yue-Hua Li, Jue-Jin Wang, Yu-Ming Kang and Guo-Qing Zhu made contributions to the conception and design of the study. Tong-Yan Liu and Jue-Jin Wang performed the analysis. Tong-Yan Liu and Guo-Qing Zhu wrote the manuscript. All authors contributed to the discussion and revision of the manuscript.
We give special thanks for the generous support of the Collaborative Innovation Center for Cardiovascular Disease Translational Medicine.
This study was supported by the National Natural Science Foundation of China [grant numbers 31171095, 31271213 and 91439120], and a project funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD).
body mass index
Dulbecco’s modified Eagle’s medium
fibronectin type III domain-containing protein 5
forkhead box transcription factor O1
glycogen synthase kinase-3
glucose tolerance test
insulin tolerance test
white adipose tissue