The role of adipokines in skeletal muscle inflammation and insulin sensitivity

The adipokines leptin, adiponectin, resistin and visfatin are known mediators of inflammation and have all been implicated in metabolic diseases, including T2D. Below, we highlight studies conducted on these adipokines which relate to their functional role in skeletal muscle insulin signalling, and summarise these findings in Table 1 in relation to animal and human data.

The role of leptin as an inflammatory adipokine in metabolic disorders is well studied. Systemic levels of leptin positively correlate with both BMI and waist circumference, and are associated with the development of insulin resistance [25, 26]. Several studies have reported that leptin impacts on skeletal muscle insulin signalling. Stimulation of the rat L6 skeletal muscle cell line with recombinant leptin reduced phosphorylation of the insulin receptor substrate-1 (IRS-1) and impaired glucose uptake, suggesting that leptin promotes insulin resistance [27]. However, in contrast, leptin stimulation of murine C2C12 myotubes was found to increase glucose uptake, whilst overexpression of leptin in a ‘skinny’ mouse model increased insulin sensitivity [28, 29]. These contrasting data highlight the need to conduct functional studies on leptin in human myotubes. To this end, Yau et al. reported that leptin increased AKT phosphorylation in commercially available human myotubes [30]. However, to date the functional role of leptin on human skeletal muscle insulin signalling is still understudied.

Adiponectin is considered to be a beneficial adipokine in relation to metabolism; plasma concentrations inversely correlate with weight, central obesity, risk of T2D and insulin resistance in humans [31, 32]. Furthermore, maintenance of a low calorie intake increases both adipocyte expression of adiponectin and circulatory concentrations [33].

Three different molecular weight isoforms of adiponectin are found in the circulation, of which the high molecular-weight isoform is beloved to be the most functional in terms of glucose homeostasis. Functional studies suggest that adiponectin promotes insulin sensitivity in skeletal muscle. In C2C12 myocytes, adiponectin increases fatty acid oxidation via sequential activation of AMPK, p38 MAPK and PPARα [34, 35] and promotes glucose uptake [35]. Similarity, in L6 myotubes adiponectin induces glucose transporter 4 (GLUT4) translocation and glucose uptake [36]. In vivo, adiponectin knockout mice demonstrate an obese, insulin resistant phenotype, whereas systemic administration of adiponectin, or its delivery as a transgene direct to skeletal muscle, improves insulin sensitivity [37‐40]. Adiponectin has also been shown to induce fat oxidation via AMPK activation in human myotubes, and further, this mechanism was found to be impaired in myotubes from obese T2D patients [41]. Critically, this suggests that the function of adiponectin as a promoter of insulin sensitivity translates to humans. Furthermore, it suggests that impairment of adiponectin function in skeletal muscle of obese T2D patients may contribute in the development of insulin resistance.

First identified in murine adipocytes as a secreted protein capable of inducing insulin resistance [42], resistin is a pro-inflammatory adipokine that induces the secretion of TNFα and IL-6 from various cell types including PBMCs and pancreatic acinar cells [43, 44].

The correlation between plasma resistin with both obesity and insulin resistance in humans support a role for resistin in the development of insulin resistance [45, 46]. In vitro, studies have demonstrated a reduction in AKT phosphorylation and glucose uptake in C2C12 and L6 myotubes stimulated with recombinant resistin [47‐49]. However, at present few studies have investigated the functional role of resistin in the development of insulin resistance in human skeletal muscle cells.

Visfatin (also called NAMPT) is termed an ‘adipokine-enzyme’ due to the NAD biosynthesis function of its intracellular form (iVisfatin/iNAMPT), also exists extracellularly (eVisfatin/eNAMPT) [50, 51]. eVisfatin is primarily produced and secreted from visceral adipose tissue, where it is more highly expressed in obese individuals [52]. Similarly, higher systemic levels of evisfatin are associated with obesity, ageing and the development of T2D [53‐55].

Regarding the role of visfatin in mediating insulin sensitivity, overexpression of visfatin in male wistar rats increased whole body insulin sensitivity [56], and in adipose tissue and liver, promoted insulin-mediated IRS-1 phosphorylation [56]. Data on the function of visfatin in skeletal muscle insulin sensitivity is limited to studies in rodents. Visfatin increases glucose transport in rat skeletal muscle fibres [57]. Furthermore, in C2C12 myotubes, visfatin activates AMPK/p38 MAPK, induces GLUT4 expression and translocation, and promotes glucose uptake [57]. Based on these data, similar insulin sensitizing effects may occur in human skeletal muscle.

In addition to the well-known adipokines, proteomic studies of adipose tissue have identified several less characterised adipokines that may also play important roles in mediating skeletal muscle insulin sensitivity. At present the functional effects of the majority of these novel adipokines on human skeletal muscle insulin sensitivity s poorly understood. Some of the more prominent novel adipokines are discussed below and also summarised in Table 2.

First reported as a 47KDa protein in the visceral adipose tissue of the genetically obese OLETF rats [58], administration of vaspin to obese mice increased insulin sensitivity and glucose tolerance [59]. Additionally, subcutaneous adipose tissue expression of leptin, resistin and TNFα was suppressed, whilst GLUT4 and adiponectin expression was increased following vaspin administration [59]. Similar increases in insulin sensitivity have since been reported in db/db and C57BL6 mice following recombinant vaspin delivery [60]. Central administration of vaspin to obese mice resulted in a sustained suppression of appetite that resulted in reduced bodyweight and plasma glucose concentrations [61]. Furthermore, transgenic mice overexpressing vaspin displayed improved glucose tolerance, reduced systemic IL-6 concentrations and were protected from obesity when fed a high fat diet [62].

In humans, vaspin expression has been reported in several tissues including subcutaneous adipose tissue, skin, stomach and skeletal muscle [61, 63, 64]. Serum concentrations of vaspin in non-diabetic and diabetic patients positively correlate with BMI, bodyweight and impaired glucose tolerance [65‐68]. Given the functional effects of vaspin demonstrated in the rodent models, its increased expression with BMI in humans may reflect a compensatory mechanism.

The effect of vaspin on insulin signalling and metabolism in human skeletal muscle is currently undetermined. Similarly, the mechanism of action and receptor for vaspin has also not been elucidated. Recently, it was reported that in HepG2 cells vaspin binds glucose-regulated protein (GRP78), a 7KDa voltage-dependent anion channel. Further, stimulation of H-4-II-E-C3 cells with recombinant vaspin activated AKT and AMPK signalling pathways, which was prevented by GRP78 inhibition [62]. Vaspin may therefore mediate its effects on insulin signalling via binding to GRP78. However, at present the expression of GRP78 has not been profiled in human adipose or skeletal muscle tissue, nor the functional studies conducted in human skeletal cells to validate GRP78 as the vaspin receptor.

FGF-21 is established as a key mediator of fat oxidation and in energy homeostasis [69‐71]. Numerous studies report that serum concentrations of FGF-21 are elevated in obese individuals and positively correlate with insulin resistance, BMI, % fat mass and circulatory concentrations of leptin and LDL [72‐75]. Although predominantly produced by the liver, FGF-21 is also expressed in adipose tissue, where it is more highly expressed in both obese and diabetic mouse models.

In vivo, administration of FGF-21 to mice fed a high fat diet decreased intramuscular triglyceride content, increased insulin sensitivity and glucose uptake, and elevated secretion of adiponectin from adipocytes [76]. Continuous cerebral administration of FGF-21 for 2 weeks increased whole body insulin sensitivity in rats with dietary-induced obesity [77], whilst daily intravenous or subcutaneous delivery of FGF-21 for 6 weeks improved glucose handling in diabetic rhesus monkeys [78]. Following such positive effects on insulin sensitivity and glucose tolerance, two FGF-21 mimetics (LY2405319 and PF-05231023) have progressed to phase 1 clinical trials (NCT01869959, NCT01923389) [79‐83], and antibodies targeting FGFR1c/b-Klotho have been developed [84, 85].

With regards to a direct functional role of FGF-21 in skeletal muscle, incubation of isolated mouse EDL muscle with FGF-21 increased insulin-stimulated glucose uptake, and in human myotubes FGF-21 increased both basal and insulin-stimulated glucose uptake [86]. Furthermore, FGF-21 has also shown to prevent palmitate-induced insulin resistance in primary human myotubes by inhibiting stress kinases and NF-κB [87].

Chemerin, was initially described as a novel chemoattractant for macrophages and dendritic cells via activation of several GPCRs including CMKLR1/ChemR23, GPR1, and CCRL2 [88, 89]. More recent data suggests chemerin plays an important role in the differentiation of human adipocytes [90, 91], and in the development of insulin resistance. Circulatory concentrations of chemerin are associated with obesity, diabetes and metabolic syndrome [92‐94] . Furthermore, adipose tissue from obese subjects exhibits greater secretion of chemerin [95].

At present, in vivo studies have drawn differing conclusions regarding the role of chemerin in the development of insulin resistance. Becker et al. reported that overexpression of chemerin increased insulin resistance in LDL-receptor deficient mice fed a high fat diet, as evidenced by reduced insulin-mediated AKT phosphorylation [96]. Importantly this effect was only observed in skeletal muscle, and not liver or pancreas [96]. Additionally, glucose handling and serum insulin concentrations were reduced by chemerin administration to both obese and diabetic mice [97]. However, no such effect was observed following chemerin administration to control mice. In contrast, Takahashi et al. showed that chemerin knockout mice display increased skeletal muscle insulin resistance, due to a disruption of hepatic glucose production and reduced insulin secretion from pancreatic Beta cells [98]. Additionally, transgenic mice overexpressing chemerin were reported to have increased skeletal muscle insulin sensitivity [98].

In vitro studies provide support for chemerin as a driver of insulin resistance. Pre-treatment of C2C12 myotubes with chemerin reduced insulin-stimulated glucose uptake, while increasing the secretion of pro-inflammatory cytokines including IL-6 and TNF-α [99]. Additionally, treatment of primary human myotubes with recombinant chemerin reduced insulin-stimulated glucose uptake [95]. Further cross-talk studies with primary human myocytes and myotubes, particularly from obese and diabetic cohorts may help to clarify the function of chemerin in human metabolic disease states.

Preadipocyte factor 1 (Pref-1) is a transmembrane protein processed to generate a circulating form, which is also known as Foetal Antigen 1 (FA1) [100]. Studies have described an association of increased Pref-1/FA1 serum concentrations with obesity and T2D [101, 102]. Pref-1 is also reported to negatively regulate adipogenesis, with Pref-1 deficient mice displaying significant obesity and stunted growth [103, 104]. Overexpression of Pref-1 in mice promotes a lipodystrophic phenotype and insulin resistance via decreased skeletal muscle glucose uptake and impaired skeletal muscle insulin signalling [105].

In humans, Pref-1 stimulation of myotubes from lean, obese, and T2D patients with did not affect insulin sensitivity. However Pref-1 did induce the production of the pro-inflammatory IL-6 and CCL2 [106], and thus chronic exposure of muscle to pathological levels of Pref-1 may impair insulin sensitivity indirectly. Clearly, further studies utilising human myotubes are warranted to fully determine the functional role of Pref-1 in skeletal muscle insulin sensitivity.

Follistatin-like 1 (FSTL1) is a glycoprotein with homology to osteonectin and its expression is associated with systemic inflammatory diseases including rheumatoid arthritis, lupus and ulcerative colitis. Several in vitro studies have established FSTL1 as a pro-inflammatory cytokine. For example, over-expression of FSTL1 in the fibroblast-like COS7 cell line or in human U937 monocytes induced the secretion of pro-inflammatory cytokines IL-6, TNF-α and IL-1β [107].

With regards to adipose biology, FSTL1 is highly expressed in 3 T3-L1 pre-adipocytes and is implicated in their differentiation [108, 109]. Furthermore, stimulation of 3 T3-L1 adipocytes with recombinant FSTL1 inhibited insulin signalling [108] In vivo, increased adipose tissue expression of FSTL1 is reported in the leptin-deficient ob/ob mouse, and in humans serum levels of FSTL1 positively correlate with BMI [108]. Despite being expressed and secreted by human myotubes [110] no studies to date have reported the functional effects of FSTL1 on skeletal muscle insulin signalling, using either rodent or human cells.

SPARC (osteonectin) was first discovered as a glycoprotein secreted from bone. However, it is now known that SPARC is also expressed and secreted from adipose tissue. SPARC adipose tissue expression is increased in dietary-induced obesity in rats [111]. In humans, SPARC is secreted from adipose tissue, implicated in adipocyte differentiation and hyperplasia [112], and its expression in adipose tissue correlates with fat mass [113]. Furthermore, serum levels of SPARC are associated with insulin resistance, dyslipidemia and inflammation in patients with gestational diabetes mellitus (GDM) [114]. Mechanistically, overexpression of SPARC in 3 T3-L1 adipocytes downregulated GLUT4 expression and inhibited insulin-stimulated glucose uptake [111]. Given these data, it seems likely that SPARC would impair skeletal muscle insulin signalling. At present, these studies have not yet been conducted and so its functional role in skeletal muscle is not established.

CTRP3 is a member of a family of proteins which includes adiponectin. Similarly to adiponectin, CTRP3 has been identified as an anti-inflammatory adipokine. In humans, CTRP3 levels in the serum are lower in obese subjects compared to normal-weight individuals [115, 116], and negatively correlate with markers of insulin resistance [116]. In vitro, CTRP3 inhibits LPS-induced expression of pro-inflammatory cytokines in human macrophages [117], whilst RNAi-mediated knockdown in preadipocytes increased the expression of chemokines and reduced adiponectin expression [118]. Its functional role in skeletal muscle insulin signalling has not been characterised.

Originally identified as a lectin-binding protein [119], Omentin-1 (intelectin-1) is highly expressed in visceral adipose tissue [120]. In humans, systemic concentrations and adipose tissue expression of Omentin-1 are lower in obese individuals [120] and negatively correlate with BMI and insulin resistance [120]. Furthermore, lower serum levels of omentin-1 are observed in newly diagnosed T2D patients and its secretion from human adipose tissue is decreased by both insulin and glucose [121, 122].

In vitro, studies support a role for Omentin-1 as an anti-inflammatory adipokine, which suppresses the activity of TNF-α in vascular inflammation via inhibiting p38 and JNK pathways. A role for Omentin-1 in promoting insulin sensitivity is supported by studies in human adipocytes where recombinant Omentin-1 induced AKT phosphorylation and enhanced insulin-stimulated glucose uptake [123]. Thus far, studies to determine its functional role in skeletal muscle using either rodent models or human tissue have not been reported.

Lipocalins are a functionally diverse group of proteins with a highly conserved tertiary structure that have been implicated in inflammation and immune responses. Importantly, a number of lipocalins, most notably lipocalin-2 (LCN2) and RBP4 have been associated with adipose tissue expression and obesity. Recently, a new member of the lipocalin family was identified, termed lipocalin-14 (LNC14), which in mice was found to be predominantly expressed in WAT and was downregulated in dietary-induced obese mice [124]. Furthermore, adenovirus over-expression of LCN14 in obese mice improved insulin sensitivity [124].