Irisin Stimulates Muscle Growth-Related Genes and Regulates Adipocyte Differentiation and Metabolism in Humans

ORIGINAL ARTICLE Irisin stimulates muscle growth-related genes and regulates adipocyte differentiation and metabolism in humans

JY Huh1, F Dincer1, E Mesfum1 and CS Mantzoros1,2

BACKGROUND: Irisin is a recently identified exercise-induced myokine suggested to induce browning of white adipocytes. Deficiency of myostatin, and thus stimulation of muscle growth, has also been reported to induce irisin and its precursor FNDC5 expression in muscle and drive the browning of white adipocytes in mice, implying that irisin may be related to muscle growth in addition to its beneficial effects in adipocytes. In humans, the effect of irisin in muscle hypertrophy as well as adipocyte metabolism has not been fully investigated. METHODS: Primary cultured human myocytes/adipocytes and 3T3-L1 cells were used to examine irisin-regulated gene/protein expression. Lipid accumulation, ATP content, glycolysis, lipolysis and metabolite profile were measured in control and irisin-treated (10 and 50 nM) adipocytes. RESULTS: In human myocytes, FNDC5 mRNA and irisin secretion were increased during myogenic differentiation, along with PGC1α and myogenin expression. Irisin treatment significantly increased insulin-like growth factor 1 and decreased myostatin gene expression through ERK pathway. PGC1α4, a newly discovered PGC1α isoform specifically related to muscle hypertrophy, was also upregulated. In human adipocytes, irisin induced uncoupling protein 1 and consequently increased adipocyte energy expenditure, expression of metabolic enzymes and metabolite intermediates, resulting in inhibition of lipid accumulation. Irisin and FNDC5 treatment also reduced preadipocyte differentiation, suggesting an additional mechanism in suppressing fat mass. CONCLUSIONS: These results suggest that irisin/FNDC5 has a pleiotropic role in muscle and improvement of adipocyte metabolism in humans. International Journal of Obesity (2014) 38, 1538–1544; doi:10.1038/ijo.2014.42 Keywords: irisin; myokine; muscle hypertrophy; adipocyte browning

INTRODUCTION genes via p38 MAPK and ERK signaling pathway.13 Although the Physical inactivity results in increased risk of type 2 diabetes and browning effect of irisin has been reported in mouse adipocytes, it cardiovascular disease,1,2 and therefore exercise and increased remains to be shown whether these effects can be translated to energy expenditure are regarded as an effective therapeutic humans. More importantly, besides browning gene expression approach.3,4 The complex mechanisms underlying the benefits and UCP1 induction, the functional outcome of irisin treatment on of exercise can be explained either by local effect in muscle, energy expenditure and metabolic function of human adipocytes including muscle growth and enhanced glucose/lipid remains to be elucidated. 14 metabolism,5 or by endocrine effect through exercise-induced The fact that irisin is predominantly expressed in muscle raises cytokines, classified as ‘myokines’.6 Discovery of myokines has a hypothesis that irisin, like other myokines such as interleukin-6 helped to understand the crosstalk between muscle and other (IL-6), IL-15, myonectin, etc.,15 could influence muscle cell metabolic organs such as adipose tissue and liver, but the impact metabolism. In the initial report by Bostrom et al., no evident of these myokines are not completely understood. effect of irisin on muscle gene expression was observed.7 Irisin is a recently discovered myokine suggested to mediate Intriguingly, recent experimental evidence in mice demonstrated the beneficial effects of exercise.7 There is limited evidence on the that deficiency of myostatin, a myokine that has an important role physiology of irisin, but it has been suggested that peroxisome in negative regulation of muscle growth and development, leads proliferator-activated receptor (PPAR)-γ coactivator-1α (PGC1α)is to PGC1α activation and irisin secretion in muscle,16 resulting in an upstream regulator of irisin and induces browning in browning of white adipocytes. Also, we have previously observed subcutaneous adipose tissue through upregulation of uncoupling a positive correlation between circulating insulin-like growth protein 1 (UCP1).7 In mice, overexpression of the irisin precursor factor 1 (IGF1) and irisin.14 Given the importance of IGF1- fibronectin type III domain-containing protein 5 (FNDC5) resulted myostatin system in exercise-induced muscle growth,17 there is in improvement of diet-induced insulin resistance.7 Therefore, a possibility that irisin could be involved in exercise-induced irisin is an attractive target for the treatment of obesity and its muscle hypertrophy by acting directly on muscle cells. related metabolic disorders.8–12 In another report, it was found in Here, we examined the regulation of irisin/FNDC5 during 3T3-L1 mouse adipocytes and rat primary adipocytes that myogenic differentiation and the effect of irisin on muscle recombinant irisin treatment upregulated browning-specific hypertrophy in human skeletal muscle cells (HSMCs). In addition,

1Division of Endocrinology, Diabetes and Metabolism, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA and 2Section of Endocrinology, Boston VA Healthcare System, Harvard Medical School, Boston, MA, USA. Correspondence: Dr CS Mantzoros, Division of Endocrinology, Diabetes and Metabolism, Beth Israel Deaconess Medical Center, Harvard Medical School, 330 Brookline Avenue, Feldberg 876, Boston, MA 02215, USA. E-mail: [email protected] Received 24 December 2013; revised 7 February 2014; accepted 4 March 2014; accepted article preview online 11 March 2014; advance online publication, 8 April 2014 Irisin in muscle and adipose tissue metabolism JY Huh et al 1539 we expand on the previous observation in mouse to human Biochemical measurements adipocytes to evaluate the effect of exogenous irisin on primary ATP was measured in cell lysates with a bioluminescence kit (Promega, human preadipocyte differentiation and mature adipocyte Madison, WI, USA). Glycolysis and lipolysis were measured by release of function. Our data show that irisin treatment is able to not only lactate and free fatty acid/glycerol in the media, respectively (Cayman regulate adipocyte metabolism but also muscle growth, both of Chemical, Ann Arbor, MI, USA, and Zen-Bio, Research Triangle Park, which could potentially be used as a target for prevention and NC, USA). Irisin and IL-6 secretion in the culture media of skeletal muscle cells were measured with commercially available ELISAs (Phoenix treatment of obesity, related metabolic diseases and muscle Pharmaceuticals, Burlingame, CA, USA; R&D Systems, Minneapolis, MN, disorders. USA, respectively).

MATERIALS AND METHODS Metabolomics Human adipocyte samples were prepared as previously described21 and Primary human skeletal muscle cell culture analyzed in the BIDMC Mass Spectrometry Core Facility. MetaboAnalyst22 Thigh muscle (vastus lateralis) was collected from obese but apparently was used to discover a biologically meaningful pattern (http://www. − healthy subjects (age 41.0 ± 7.9 years, BMI 43.5 ± 1.7 kg m 2) and cultured metaboanalyst.ca). For metabolite set enrichment analysis, data were as previously described.18 The morphology and growth of the isolated mapped according to the Human Metabolome Database (http://www. myocytes were normal and their characteristics were maintained until hmdb.ca), and the Metabolic Pathway Library (currently 88 entries) was passage 7. During myogenic differentiation, media was changed every 24 h chosen to assess the data using the global test package.23 and the media was collected on day 0, 2, 5, 8, 12. For inhibition experiments, skeletal muscle cells were incubated with ERK inhibitor U0126 (10 μM, Cell Signaling, Danvers, MA, USA) 2 h before irisin treatment. Statistics SPSS software 19.0 (SPSS Inc., Chicago, IL, USA) were used and data are shown as mean ± s.e. unless stated otherwise. Mean values obtained from Primary human adipocyte culture multiple experiments were compared by ANOVA with subsequent Fisher’s Subcutaneous human adipocytes were obtained from the same patients as signiﬁcant difference method. Metabolite changes were analyzed by above. Primary adipocyte culture was performed as described previously.19 Student’s t-test. P values of o0.05 were considered as statistically To examine the effect on adipocyte differentiation, the cells were treated signiﬁcant for all analyses. with 10 or 50 nM recombinant irisin (Aviscera Bioscience, Santa Clara, CA, USA) every 2 days, starting from 2 days before differentiation.

RESULTS 3T3-L1 adipocyte culture Irisin/FNDC5 is increased during myogenic differentiation 3T3-L1 preadipocytes were cultured as previously described.20 During the first 2 days of differentiation, the medium was supplemented with 1 μM In primary cultured human cells, myotube FNDC5 mRNA level was − 1 dexamethasone, 1 μgml insulin, and either 0.5 mM IBMX or 1 μM approximately 500 times higher compared with stromal vascular rosiglitazone. Recombinant irisin and FNDC5 (Aviscera Bioscience) were cells or adipocytes (Figure 1a). Also, the FNDC5 mRNA level in used to compare the effect on preadipocyte differentiation. mature adipocytes was significantly lower than in stromal vascular cells. As myocytes undergo cell differentiation to become myotubes, the change in FNDC5 expression was monitored Gene expression analysis during myogenic differentiation. FNDC5 mRNA level was drama- Total RNA was extracted from adipocytes using Trizol (Invitrogen, Carlsbad, tically upregulated at early time points during differentiation, CA, USA) according to a standard protocol. mRNA levels were measured by 14 peaking at day 5 and reaching a plateau from day 8 onward real-time PCR using TaqMan Gene Expression Assays as described. (Figure 1b). This expression pattern was similar to those of myogenin and PGC1α, although the magnitude of FNDC5 increase Western blot analysis was smaller, whereas another myokine IL-6 had no related pattern. Western blot analysis in adipocytes was performed as previously In contrast to the early increase in FNDC5 mRNA levels, irisin described.18 The membranes were incubated with primary antibody (aP2 secretion in the media was not increased until day 8 (Figure 1c). and PPARγ from Cell Signaling; FAS, PGC1α, ATGL, Akt, ERK1/2 and β-actin Following the pattern of gene expression level, IL-6 secretion was from Santa Cruz Biotechnology, Dallas, TX, USA; UCP1 from Abcam, decreased until day 5 and significantly increased again at day 8. Cambridge, MA, USA) overnight. Basal irisin levels were higher than IL-6 in the media.

Figure 1. Regulation of irisin in human skeletal muscle cells. (a) Real-time PCR was performed in primary cultured human myotubes, stromal vascular cells and mature adipocytes. (b) mRNA levels of FNDC5, PGC1α and IL-6 were measured during myogenic differentiation. Myogenin was measured as an index of differentiation. (c) Irisin and IL-6 secretion in the media for 24 h in each time point was measured by ELISA. Values are means ± s.e. of three to four experiments. *Po0.05 vs day 0, †Po0.05 vs day 5.

Figure 2. Effect of irisin on genes related to muscle hypertrophy in primary human skeletal muscle cells via ERK. (a) Differentiated primary human skeletal muscle cells were incubated with 0, 10 or 50 nM irisin for 6 h. Gene expression levels were measured by real-time PCR. Values are means ± s.e. of four individual experiments. (b and c) Differentiated primary human skeletal muscle cells were incubated with 50 nM irisin for indicated time points. The phosphorylation of ERK and Akt was measured by western blot analysis. (d) Cells were incubated with 10 μM U0126 (ERK inhibitor) for 2 h before 50 nM irisin treatment for 6 h. Values are means ± s.e. of four individual experiments. *Po0.05 vs control (no irisin treatment), †Po0.05 vs irisin-treated cells.

Irisin regulates genes related to muscle hypertrophy First, to investigate the effect of irisin on preadipocyte differentia- To examine the hypothesis that irisin could regulate muscle tion, stromal vascular cells from human subcutaneous adipose growth, HSMCs were treated with low and high physiological tissue were treated with low (10 nM) or high (50 nM) physiological doses of irisin (0, 10 and 50 nM irisin) for different time periods concentrations of recombinant irisin throughout the differentia- (2, 6 and 24 h). Gene expression analysis showed that at 6 and 24 h tion period. As a result, irisin inhibited the lipid accumulation after treatment, irisin dose-dependently increased IGF1 and (Figure 3a), along with significantly reduced gene/protein expres- decreased myostatin mRNA levels (Figure 2a and Supplementary sions of adipocyte protein 2 (aP2), PPARγ and fatty acid synthase Figure 1), the two main factors for muscle growth. This was (FAS; Figures 3b–f). In mouse adipocytes, similar results were accompanied by upregulation of PGC1α4, their upstream reg- observed (Supplementary Figure 2). Comparing the effects of irisin ulator, confirming the positive effect of irisin treatment on muscle and FNDC5 on mouse adipocytes showed that irisin exerted a growth. It is interesting to note that 2 h treatment of irisin more potent effect on the inhibition of differentiation. In addition, significantly upregulated FNDC5 expression (Supplementary the reduced protein expression of aP2 and PPARγ, and gene Figure 1), but followed by significant reduction of FNDC5 and its expression of aP2, CCAAT/enhancer-binding protein α (C/EBPα), upstream mediator, PGC1α, at 6 and 24 h. PPARγ and adiponectin by both irisin and FNDC5 were partially reversed by rosiglitazone supplementation during differentiation, suggesting that suppression of preadipocyte differentiation by Irisin regulates muscle growth through ERK pathway irisin/FNDC5 is influenced, at least in part, via PPARγ-dependent Signaling studies in adipocytes have showed that the browning mechanisms (Supplementary Figure 2). effect of irisin was mediated through p38 MAPK and ERK but not Akt pathway.13 In HSMCs, we observed that, similar to adipocytes, ERK was significantly phosphorylated as early as 10 min after irisin Irisin alters metabolic gene expression in human adipocytes stimulation (Figure 2b), whereas phosphorylation of Akt was FNDC5 treatment has been reported to induce ‘browning’ genes unaltered (Figure 2c). To confirm whether the growth-related gene in mouse adipocytes.7 To expand, mature human adipocytes were induction was mediated through ERK, HSMCs were pretreated with treated with irisin in a time course of 8 days, and genes regulating ERK inhibitor (U0126). As a result, ERK inhibitor significantly reversed adipocyte function were examined. The adipocytes were fully the induction of PGC1α4 and IGF1 gene expression (Figure 2d). differentiated before treatment to exclude the effect of irisin on preadipocyte differentiation. As shown in Figure 4a, browning genes, including UCP1, PRDM16 and CIDEA, were induced Irisin inhibits adipocyte differentiation in human and mouse the most after 8 days of treatment, and ELDVL3 after 2 days adipocytes of treatment. Expression of adipokines and mitochondrial In addition to the novel role of irisin on muscle growth, we next biogenesis-related genes was also increased by 4 and/or 8 days sought to examine the role of irisin on adipocyte metabolism. of irisin treatment (Figures 4b and c). Interestingly, genes

Figure 3. Effect of irisin on human preadipocyte differentiation. (a) Oil Red O staining of human adipocytes at 14 and 28 days after differentiation, treated with irisin. (b and c) Real-time PCR results for in human adipocytes treated with 50 nM irisin at 14 and 28 days after differentiation. (d) Western blots against adipocyte protein 2 (aP2) and fatty acid synthase (FAS) at 28 days after differentiation in human adipocytes treated with irisin. β-Actin was used as an internal control. (e and f) Relative quantification of western blots. Values are means ± s.e. of four experiments. *Po0.05 vs control. encoding proteins involved in glucose and lipid metabolism, the tricarboxylic acid cycle, namely citric acid, cis-aconitic acid and including GLUT4, CPT1a, PPARα and HSL, were also significantly isocitric acid, were elevated by irisin treatment. Metabolite set increased (Figure 4d). enrichment analysis showed that pathways related to glucose and lipid metabolism were largely enhanced (Figure 6b). Amino acid Irisin affects mature human adipocyte metabolism pathways were also affected by irisin treatment. The metabolome view confirmed that irisin causes profounding effects in adipocyte In line with the transcriptional changes, irisin treatment induced metabolism and therefore coincide with the functional data observed. UCP1 protein expression in human adipocytes (Figure 5a). As a result, intracellular ATP levels were depleted by irisin treatment, comparable to the effect of the oxidative phosphorylation DISCUSSION uncoupler FCCP (Figure 5d). Morphological and colorimetric The discovery of irisin has pointed to a novel pathway for energy analyses revealed that irisin-treated adipocytes were smaller homeostasis and opened an opportunity for development of (Figure 5b) with less amount of lipid accumulation (Figure 5c). novel pharmaceutical compounds to treat metabolic diseases. Because of limited oxidative respiration by uncoupling in the Although a mouse study has revealed the beneficial role of irisin mitochondria, the glycolytic pathway was significantly induced in on systemic metabolism, which type of tissue or organs are cells treated with 50 nM irisin, as determined by the secretion of affected by irisin and what the underlying mechanisms are remain lactate in the media (Figure 5e). Analysis of free fatty acid and largely unknown. Here, we provided evidence for the pleiotropic glycerol secretion in the media revealed that irisin not only effect of irisin on not only adipocyte but also myocyte inhibited basal lipolysis but also isoproterenol-induced lipolysis in metabolism. The main findings are that (i) irisin secretion and human adipocytes (Figures 5f and g). In view of elevated lipid FNDC5 mRNA expression is upregulated during myocyte differ- utilization by browning, a reduction in lipolytic rate may seem entiation, (ii) irisin induces expression of PGC1α4 and IGF1, and unlikely. However, regarding the overall low lipid content of irisin- represses myostatin gene expression through ERK pathway, treated cells, adipocytes may have had lower output of lipids. (iii) irisin not only shifts the gene expression and metabolic Moreover, western blot results showed that adipose triglyceride profile of mature human adipocytes toward increased energy lipase (ATGL) was increased, whereas FAS was decreased by irisin expenditure but also inhibits differentiation of preadipocytes. treatment (Figure 5a), implying that through induction of UCP1, Therefore, irisin exerts different benefits in various cell types, irisin breaks down stored fat to be used for intracellular which could benefit not only obesity-related metabolic diseases metabolism and at the same time inhibits the synthesis of lipids. but also muscle disorders. Our results from human adipocytes clearly show at transcrip- Irisin alters metabolite profile of human adipocytes tional, translational and metabolite level that irisin can induce To add another dimension, changes in metabolites were analyzed. browning and regulate energy homeostasis. Along with UCP1 Relative changes in metabolites by irisin treatment are shown in induction and depletion of intracellular ATP levels, irisin turns on Figure 6a. It is interesting to note that acetoacetyl-CoA, a product the mechanism to enhance energy circulation in adipocytes, as of fatty acid degradation, and some of the key intermediates of evidenced by upregulation in intermediates of citric acid cycle and

Figure 4. Changes in gene expression by irisin treatment in human adipocytes. (a–d) Changes in various gene expression related to adipocyte metabolism at 1, 2, 4 and 8 days of 10 nM irisin treatment. Gene expression was normalized by the corresponding untreated controls at each time point. Values are means ± s.e. of four experiments. *Po0.05 vs control.

lactate production. Increased ATGL may indicate that irisin is able Myokines are key molecules in crosstalk between muscle and to induce ‘melt down’ of stored fat but on the otherhand, reduced other metabolic organs, but many of the myokines are also known extracellular levels of fatty acids and glycerol by irisin treatment to exert autocrine/paracrine effect. For example, IL-15, which is may imply that irisin is able to preserve adipocytes from ‘leaking’ also mainly expressed in muscle, has been reported to inhibit lipid deleterious lipids into other peripheral organs. It remains accumulation and induce adiponectin secretion in adipocytes and uncertain exactly how irisin regulates adipocyte metabolism, also induce glucose/insulin sensitivity in muscle.6,30 Likewise, but the previous study suggested p38 MAPK and ERK to be the independent from the results on adipocyte browning, we provide signaling target of irisin in mouse/rat adipocytes. Whether evidence for the first time on irisin-induced muscle growth the same signaling pathway could be applied in humans remains through increased IGF1 and decreased myostatin levels in a dose- to be further studied in detail. dependent manner. Ruas et al.31 have recently shown that Although a previous study has found that irisin has no effect on resistance exercise induces a novel isoform of PGC1α, PGC1α4, mouse preadipocyte differentiation,7 we have discovered a to mediate muscle hypertrophy. In line with this observation, significant inhibitory effect of irisin on both human and mouse PGC1α4 mRNA level was significantly increased by irisin treatment preadipocyte differentiation. This is somewhat surprising, as and regulated the downstream molecules, IGF1 and myostatin. factors known to induce browning, such as bone morphogenetic IGF1 is a positive regulator and myostatin is a negative regulator protein 7, have been reported to accelerate preadipocyte of muscle growth17 and thus regulation of these genes by irisin differentiation.24 However, given that irisin promotes energy imply that irisin could be mediating the effect of exercise that expenditure in mature adipocytes and inhibits lipid accumulation, primarily act on enhancing muscle mass and strength. Study in it seems reasonable that preadipocytes are prohibited from myostatin knockout mice has led us to think that irisin could be accumulating lipids. This may be of value as dual role of irisin in related to IGF1-myostatin system and moreover, to muscle anti-obesity effect, by inhibiting adipocyte formation and increas- hypertrophy. Although the previous study suggested that lack of ing metabolic rate in mature adipocytes. myostatin induces FNDC5 expression,16 our data claim that Studies have implicated the possibility of irisin production from increased irisin could inhibit myostatin, implying a bilateral effect. tissues other than skeletal muscle, such as white adipose Studies in humans so far have not observed a significant increase tissue,25,26 cardiac muscle27 or neurons,28,29 in rodents and partly in either circulating irisin or muscle FNDC5 mRNA expression by in humans. Our results from human myocytes demonstrate that resistance exercise,32,33 but the possibility remains that the local FNDC5 gene expression level is much higher than adipocytes and concentration of irisin could be much higher post-exercise, which that FNDC5 gene expression and irisin secretion is regulated would enable the autocrine/paracrine effect of irisin. The during myocyte differentiation in humans. Further study is needed transcriptional effects observed in this study should be confirmed to dissect what percentage of circulating irisin is derived from by protein expression and phenotypic changes in muscle. skeletal muscle and whether other organs are also capable of irisin Interestingly, irisin treatment resulted in upregulation of FNDC5 secretion in humans. at 2 h, whereas more long-term treatment (6 and 24 h) resulted in

Figure 5. Effect of irisin on human adipocyte metabolism. (a) Western blot analysis of human adipocytes treated with 50 nM irisin for 8 days. β-Actin was used as an internal control. (b) Oil Red O staining of human adipocytes treated with 50 nM irisin for 8 days. (c) Quantiﬁcation of lipid content by isopropanol extraction of Oil Red O. (d) ATP levels measured in lysates of human adipocytes treated with 50 nM irisin for 8 days or FCCP for 6 h. ATP concentrations were normalized to protein content and control. (e) Glycolysis was measured by the release of lactate in the media after treatment with irisin for 8 days. (f and g) Lipolysis was measured by the release of free fatty acids and glycerol in the media after treatment with irisin for 8 days. 1 μM isoproterenol was used to induce lipolysis. Open bar: no irisin, hatched bar: 10 nM irisin, full † bar: 50 nM irisin. Values are means ± s.e. of four experiments. *Po0.05 vs control, Po0.05 vs isoproterenol-treated control.

Figure 6. Effect of irisin on adipocyte metabolite proﬁle. (a) Intracellular metabolites were analyzed 4 days after treatment with 50 nM irisin. Metabolites with signiﬁcant fold changes (FCs) are shown. (b) Summary plot for Metabolite set enrichment analysis using the Metabolic Pathway Library. Metabolite sets are ranked according to P value.

© 2014 Macmillan Publishers Limited International Journal of Obesity (2014) 1538 – 1544 Irisin in muscle and adipose tissue metabolism JY Huh et al 1544 downregulation of FNDC5 and PGC1α mRNA levels. Although 7 Bostrom P, Wu J, Jedrychowski MP, Korde A, Ye L, Lo JC et al. A PGC1- direct evidence is needed, we speculate that exogenous stimula- alpha-dependent myokine that drives brown-fat-like development of white fat tion may result in a negative feedback loop to decrease and thermogenesis. Nature 2012; 481: 463–468. its endogenous expression. In the initial publication, it has 8 Kelly DP. Medicine. Irisin, light my fire. Science 2012; 336:42–43. α 9 Cunha A. Basic research: Irisin--behind the benefits of exercise. Nat Rev Endocrinol been suggested that PGC1 is an upstream regulator of irisin, 8 but now that several isoforms of PGC1α has been discovered,31 2012; : 195. 10 Polyzos SA, Kountouras J, Shields K, Mantzoros CS. Irisin: a renaissance in meta- it has to be dissected which isoforms are responsible for bolism? Metabolism: clinical and experimental 2013; 62: 1037–1044. irisin/FNDC5 regulation. 11 Bostrom PA, Fernandez-Real JM, Mantzoros C. Irisin in humans: recent advances The present study has limitations in connecting the role of irisin and questions for future research. Metab Clin Exp 2014; 63:178–180. in muscle growth and the benefits in the adipocytes. Although our 12 Park KH, Zaichenko L, Brinkoetter M, Thakkar B, Sahin-Efe A, Joung KE et al. results in vitro adds value to the therapeutic potential of irisin, Circulating Irisin in Relation to Insulin Resistance and the Metabolic Syndrome. whether irisin treatment can systematically ameliorate metabolic J Clin Endocrino Metab 2013; 98: 4899–4907. dysregulation needs to be studied in long-term in vivo studies. 13 Zhang Y, Li R, Meng Y, Li S, Donelan W, Zhao Y et al. Irisin Stimulates Browning of White Adipocytes through Mitogen-Activated Protein Kinase p38 MAP Kinase and Especially, whether irisin-induced muscle gene expression results 63 – in translational/functional change in muscle needs to be ERK MAP Kinase Signaling. Diabetes 2013; : 514 525. 14 Huh JY, Panagiotou G, Mougios V, Brinkoetter M, Vamvini MT, Schneider BE et al. examined. Assessing the direct cause-effect relationship between FNDC5 and irisin in humans: I. Predictors of circulating concentrations in serum irisin treatment and metabolic effects will largely progress when and plasma and II. mRNA expression and circulating concentrations in response genetically engineered mouse models for FNDC5 are developed. to weight loss and exercise. Metab Clin Exp 2012; 61: 1725–1738. Discovery of the irisin receptor would also largely contribute 15 Raschke S, Eckel J. Adipo-myokines: two sides of the same coin--mediators to understanding the biological function of irisin in not only of inflammation and mediators of exercise. Mediators Inflamm 2013; 2013: 320724. adipocytes but also other metabolic organs such as liver and 16 Shan T, Liang X, Bi P, Kuang S. Myostatin knockout drives browning of white pancreas. adipose tissue through activating the AMPK-PGC1alpha-Fndc5 pathway in muscle. FASEB J 2013; 27:1981–1989. In conclusion, we have explored the therapeutic potential of fi irisin on human cells and found that irisin could regulate muscle 17 Schiaf no S, Dyar KA, Ciciliot S, Blaauw B, Sandri M. Mechanisms regulating skeletal muscle growth and atrophy. FEBS J 2013; 280: 4294–4314. growth as well as adipocyte metabolism. The molecular, 18 Gouni-Berthold I, Berthold HK, Huh JY, Berman R, Spenrath N, Krone W et al. biochemical and metabolomic data on the effect of physiological Effects of lipid-lowering drugs on irisin in human subjects in vivo and in human doses of irisin have contributed in understanding the nature of skeletal muscle cells ex vivo. PloS One 2013; 8: e72858. irisin in humans. Although irisin may not replace the beneficial 19 Smih F, Rouet P, Lucas S, Mairal A, Sengenes C, Lafontan M et al. Transcriptional effects of exercise, it would still be an attractive tool for treating regulation of adipocyte hormone-sensitive lipase by glucose. Diabetes 2002; 51: metabolic diseases and muscle disorders considering its effects on 293–300. human adipocytes and myocytes. 20 Huh JY, Kim Y, Jeong J, Park J, Kim I, Huh KH et al. Peroxiredoxin 3 is a key molecule regulating adipocyte oxidative stress, mitochondrial biogenesis, and adipokine expression. Antioxid Redox Signaling 2012; 16: 229–243. 21 Yuan M, Breitkopf SB, Yang X, Asara JM. A positive/negative ion-switching, CONFLICT OF INTEREST targeted mass spectrometry-based metabolomics platform for bodily fluids, cells, The authors declare no conflict of interest. and fresh and fixed tissue. Nat Protoc 2012; 7: 872–881. 22 Xia J, Psychogios N, Young N, Wishart DS. MetaboAnalyst: a web server for metabolomic data analysis and interpretation. Nucleic Acids Res 2009; 37:W652–W660. 23 Goeman JJ, van de Geer SA, de Kort F, van Houwelingen HC. A global test for ACKNOWLEDGEMENTS groups of genes: testing association with a clinical outcome. Bioinformatics 2004; JYH researched data and wrote the manuscript. FD and EM researched data 20:93–99. and reviewed the manuscript. CM designed the studies, supervised laboratory 24 Bonet ML, Oliver P, Palou A. Pharmacological and nutritional agents promoting measurements and reviewed/edited the manuscript. browning of white adipose tissue. Biochim Biophys Acta 2013; 1831:969–985. JYH is the guarantor of this work and, as such, had full access to all the data in the 25 Moreno-Navarrete JM, Ortega F, Serrano M, Guerra E, Pardo G, Tinahones F et al. study and takes responsibility for the integrity of the data and the accuracy of the Irisin is expressed and produced by human muscle and adipose tissue in asso- data analysis. ciation with obesity and insulin resistance. J Clin Endocrinol Metab 2013; 98: This study was supported by Award Number 1I01CX000422-01A1 from the Clinical E769–E778. Science Research and Development Service of the VA Office of Research and 26 Roca-Rivada A, Castelao C, Senin LL, Landrove MO, Baltar J, Belen Crujeiras A et al. Development. FNDC5/irisin is not only a myokine but also an adipokine. PloS One 2013; 8: e60563. 27 Aydin S, Kuloglu T, Eren MN, Celik A, Yilmaz M, Kalayci M et al. Cardiac, skeletal muscle and serum irisin responses to with or without water exercise in young and old male rats: Cardiac muscle produces more irisin than skeletal muscle. REFERENCES Peptides 2013; 52C:68–73. 1 Tuomilehto J, Lindstrom J, Eriksson JG, Valle TT, Hamalainen H, Ilanne-Parikka P 28 Dun SL, Lyu RM, Chen YH, Chang JK, Luo JJ, Dun NJ et al. Irisin-immunoreactivity et al. Prevention of type 2 diabetes mellitus by changes in lifestyle among in neural and non-neural cells of the rodent. Neuroscience 2013; 240:155–162. subjects with impaired glucose tolerance. N Engl J Med 2001; 344: 1343–1350. 29 Wrann CD, White JP, Salogiannnis J, Laznik-Bogoslavski D, Wu J, Ma D et al. 2 Nocon M, Hiemann T, Muller-Riemenschneider F, Thalau F, Roll S, Willich SN et al. Exercise induces hippocampal BDNF through a PGC-1alpha/FNDC5 pathway. Association of physical activity with all-cause and cardiovascular mortality: a Cell Metab 2013; 18: 649–659. systematic review and meta-analysis. Eur J Cardiovasc Prev Rehabil 2008; 15: 30 Raschke S, Eckardt K, Bjorklund Holven K, Jensen J, Eckel J. Identification and 239–246. validation of novel contraction-regulated myokines released from primary human 3 Colberg SR, Sigal RJ, Fernhall B, Regensteiner JG, Blissmer BJ, Rubin RR et al. skeletal muscle cells. PloS One 2013; 8: e62008. Exercise and type 2 diabetes: the American College of Sports Medicine and the 31 Ruas JL, White JP, Rao RR, Kleiner S, Brannan KT, Harrison BC et al. A PGC-1alpha American Diabetes Association: joint position statement. Diabetes Care 2010; 33: isoform induced by resistance training regulates skeletal muscle hypertrophy. Cell e147–e167. 2012; 151:1319–1331. 4 Bird SR, Hawley JA. Exercise and type 2 diabetes: new prescription for an old 32 Moraes C, Leal VO, Marinho SM, Barroso SG, Rocha GS, Boaventura GT et al. problem. Maturitas 2012; 72:311–316. Resistance Exercise Training does not Affect Plasma Irisin Levels of Hemodialysis 5 Egan B, Zierath JR. Exercise metabolism and the molecular regulation of skeletal Patients. Horm Metab Res 2013; 45: 900–904. muscle adaptation. Cell Metab 2013; 17: 162–184. 33 Hecksteden A, Wegmann M, Steffen A, Kraushaar J, Morsch A, 6 Pedersen BK, Febbraio MA. Muscles, exercise and obesity: skeletal muscle as a Ruppenthal S et al. Irisin and exercise training in humans—Results from a ran- secretory organ. Nat Rev Endocrinol 2012; 8:457–465. domized controlled training trial. BMC Med 2013; 11: 235.

Supplementary Information accompanies this paper on International Journal of Obesity website (http://www.nature.com/ijo)