Introduction

The incidence of diabetes and the associated metabolic abnormalities are rapidly growing in western countries [1]. The term metabolic syndrome or syndrome X identifying insulin resistance as a pathophysiologically relevant mechanism in human disease was initially defined in 1988, and since then research has focused on identifying the molecules and pathways that are related to this contribution [2]. Today it is estimated that the prevalence of the insulin resistance syndrome in the United States [3, 4] and Europe [5] reaches 20 to 30% of the general population depending on the underlying definition [6]. Etiologically increased caloric intake and reduced energy expenditure related to modern lifestyles have been used to explain the rise in the prevalence of the metabolic syndrome [7]. The pathophysiological relevance of the increasing incidence of insulin resistance in the general population in terms of health care costs can only be estimated. Besides the “classical” manifestation of insulin resistance in diabetes, it acts as cofactor in the pathogenesis of atherosclerosis, malignant disease, acute organ failure, and other inflammation-associated diseases. From a developmental viewpoint, metabolic and immune signaling pathways share common roots and obesity-induced activation of immune responses contribute to disease development, posing an increasing challenge to health care providers.

In liver disease, insulin resistance has been recognized as an independent predictor and risk factor for the development of non-alcoholic and alcoholic steatohepatitis, chronic viral hepatitis, and hepatocellular carcinoma (HCC) [8–10]. Hyperlipidemia, insulin resistance, and conditions associated with the metabolic syndrome were also shown to profoundly modulate the response to treatment and the overall prognosis of chronic liver disease, such as chronic hepatitis C virus infection (HCV). Additionally, end stage liver disease characterized by fibrosis or cirrhosis, occurs six times more frequently in patients with insulin resistance than in patients with physiological insulin sensitivity [11–13]. Current research focuses on the delineation of the molecular mechanisms that link insulin resistance and hepatocellular injury. However, the mechanisms that promote hepatocellular injury in the context of decreased insulin sensitivity are not well defined yet. Our review summarizes insulin signaling in the context of liver disease and apoptotic hepatocellular injury focusing on signaling molecules that have been identified to mediate hepatocellular injury and decrease insulin sensitivity. The delineation of these signaling crosstalks will advance the understanding of hormones and cytokines potentially interacting with cell death and insulin signaling. Targeting insulin resistance might then substantially contribute to therapy of chronic liver disease.

Signaling pathways in diabetes and apoptosis

Insulin signaling pathways in hepatocytes

The liver is a central regulator of carbohydrate homeostasis and releases glucose according to metabolic demands. In acute or chronic liver disease the metabolic functions of the liver are disturbed and the dysregulation contributes to clinical manifestation of liver disease [14]. The failure of hepatocytes to respond to insulin in a physiological way results in uncontrolled gluconeogenesis, glycogenolysis and lipogenesis, promoting hyperglycemia, systemic insulin resistance and dyslipedemia [15, 16]. In hepatocytes, these metabolic processes are under control of insulin signaling which is initiated following binding of insulin to the insulin receptor (Fig. 1). Following ligand–receptor interaction, a class of molecules known as insulin receptor substrates (IRS) are recruited and activated by tyrosine phosphorylation [17]. The critical role of IRS in insulin responsiveness has been demonstrated through deletion of IRS-1 or IRS-2 in mice, which resulted in hepatic insulin resistance and symptoms mimicking diabetes including increased gluconeogenesis and glycogenolysis [18, 19]. Tyrosine phosphorylation of IRS results in activation of downstream effectors that mediate the metabolic and growth stimulatory effects of insulin [20]. In contrast, increased phosphorylation of IRS serine residues results in decreased tyrosine phosphorylation, and termination of the insulin signaling cascade [21]. Tyrosine phosphatases and serine kinases including c-Jun N-terminal kinase (JNK) [22], protein kinase C (PKC) [23, 24], the mammalian target of rapamycin (mTOR) [25, 26], and IκB kinase (IKK) [27] have been mechanistically implicated in the regulation of insulin resistance. Interestingly, these kinases also mediate downstream effects of cell death receptor signaling pathways (Fig. 1). Degradation of adapter molecules is a second mechanism which contributes to decreased insulin sensitivity: Oxidative stress mediated by inducible nitric-oxide synthase (iNOS) or induction of suppressors of cytokine signaling (SOCS) in response to pro-inflammatory cytokines results in insulin resistance through ubiquitination and degradation of IRS-1 [28–30]. The principle downstream effector of insulin signaling is the protein kinase Akt, as underlined by the fact that animals lacking the Akt2 isoform developed insulin resistance secondary to loss of the effects of insulin on the liver [31]. Upon activation, Akt phosphorylates and inactivates the inhibitory kinase of glycogen synthase, glycogen synthase kinase 3 (GSK3). Transcriptional effects of insulin are mediated in part by members of the winged helix/forkhead transcription factor family FoxO [32]. FoxO transcription factors are phosphorylated in response to insulin resulting in their export from the nucleus and the abrogation of their transcriptional activity. These regulatory and transcriptional effects mediate the inhibition of gluconeogenesis and glycogenolysis in hepatocytes upon insulin exposure [33] (Fig. 1). Glucose uptake in hepatocytes is achieved through transmembrane glucose transporters (GLUT)-4 which exist in preformed vesicles in the cytoplasm, and translocate to the membrane in response to insulin. Glucose uptake via GLUT-4 in hepatocytes occurs through a concentration-dependent diffusion.

Fig. 1
figure 1

Cell death- and insulin signaling pathways and crosstalks. Binding of TNF to the TNF-R1 induces the recruitment of adaptor proteins and formation of an intracellular signaling complex. The survival pathway (outlined on the right side of the TNF-R1) is activated following recruitment of TRADD, TRAF-2 and RIP. These activate the IKK complex which phosphorylates inhibitors of NF-κB, resulting in NF-κB activation. NF-κB translocates to the nucleus and activates genes that act to block TNF-induced apoptosis shown on the left side of the TNF-R1. Additionally, the TRADD/TRAF-2/RIP complex leads to activation of the JNK/c-Jun/AP-1 pathway. JNK exerts dichotomous functions and induces insulin resistance through phosphorylation of serine residues of the insulin receptor substrate (IRS)-1 molecule, while sustained AP-1 transcriptional activity in response to JNK activation triggers the apoptotic death pathway. In the death pathway, TRADD dissociates from the TNF-R1 and recruits FADD and caspases resulting in apoptosis through release of pro-apoptotic factors from mitochondria. Activation of caspases can be prevented through the caspase-8 homologue FLICE inhibitory protein (FLIP). The insulin receptor signaling pathway is activated through tyrosine phosphorylation of IRS-1 following the binding of insulin. Tyrosin phosphorylated IRS-1 leads to activation of phosphoinositide 3-kinase (PI3 K) which in turn phosphorylates Akt. Akt activates target proteins through phosphorylation and such regulates glucose uptake and gluconeogenesis. This is achieved through translocation of GLUT4 transporters to the cellular membrane, inhibition of glycogen synthase kinase 3 (GSK3) and inhibition of FoxO transcription factors. Inactivation of glycogen synthase through GSK3 and inhibition of FoxO mediated transcription of the rate controlling enzyme of gluconeogenesis phosphoenolpyruvate carboxykinase (PEPCK) results in suppressed gluconeogenesis, the physiological response to insulin. Serine kinases, including JNK1, JNK2 and the IKK complex phosphorylate IRS-1 at serine residues, preventing Akt activation and thus insulin’s inhibitory effects on gluconeogenesis and glucose uptake, leading to hyperglycemia and insulin resistance

Full size image

The importance of the above mentioned insulin signaling events in apoptosis will be pointed out in the following sections.

Apoptosis signaling pathways in hepatocytes

Liver homeostasis is achieved by a tightly regulated steady state in cell turnover involving proliferation and apoptosis of hepatocytes. A disturbance of this balance can result in apparent liver disease [34]. Compensatory activation of repair mechanisms in the liver following liver injury involves the activation of hepatic stellate cells (HSC) and promotes development of liver fibrosis, potentially resulting in chronic liver injury [35, 36]. Besides apoptosis hepatocyte cell death can also occur from necrosis or autophagy depending on the microenvironment and type of injury, however a clear differentiation between these types of cell death is not always possible [37]. Apoptosis is initiated in an autocrine or paracrine way following binding of a death ligand to its corresponding receptor, while autophagy utilizes the cells lysosomal and proteasomal system for degradation of cellular components [34, 38]. Prominent among the apoptosis receptor/ligand systems is the tumor necrosis factor receptor (TNF-R) superfamily with its most abundantly studied death receptors CD95/Apo-1/Fas and TNF-R1 (CD120a): CD95-Ligand and TNF are released from activated macrophages or expressed on cytotoxic T-cells in response to pathogens, autoimmunogens or inflammation [39–41]. The intracellular signaling complex that is recruited following ligand binding, will be exemplified for TNF in the following paragraph (see [42–44] for detailed review).

TNF exerts its biological effects by binding to either TNF receptor type 1 (TNF-R1) or TNF receptor type 2 (TNF-R2) at the plasma membrane. Following ligand–receptor interaction, an early intracellular signaling complex is formed that dissociates from the receptor–ligand complex and recruits other intracellular signaling molecules [45]. Unique for TNF signaling is the capacity to result in divergent cellular effects namely cell death or proliferation/inflammation, however the early events are common to all of the biological effects of TNF signaling [46]. In contrast to the insulin receptor, the intracellular domains of TNF-R are devoid of intrinsic kinase activity and therefore depend on homophilic protein–protein interactions between motifs for the initiation of signaling [47]. Based on the existence of one of two distinct domains, the TNF-R superfamily members are divided into two subgroups, the death domain (DD)-containing receptors and TNF-R-associated factor (TRAF) interacting receptors [48]. The apoptotic cell death pathway is activated following recruitment of the Fas-associated death domain (FADD/MORT) protein to TRADD through interaction of the death domains (DD) [49]. FADD subsequently leads to activation of caspases and the release of pro-apoptotic factors from mitochondria to sustain the apoptotic signal [50] (Fig. 1). Activation of the nuclear factor-κB (NF-κB) is triggered following TNF receptor activation through recruitment of IKK to the TNF-R signaling complex through RIP1 [51, 52]. In resting cells, NF-κB is sequestered in the cytoplasm as an inactive heterodimeric complex bound to an inhibitory counterpart IκB. Phosphorylation of IκB by IKK results in IκB degradation and dissociation from NF-κB, leading to translocation of NF-κB to the nucleus and activation of target genes [53]. In the context of TNF-induced NF-κB activation, the expression of anti-apoptotic proteins, such as Bcl-2 family members or FLIP prevents hepatocellular apoptosis through stabilizing effects on mitochondria or impaired caspase activation [54, 55] (Fig. 1). Thus, TNF-induced apoptosis can only proceed if transcriptional arrest of NF-κB target genes occurs [42, 56].

Autophagy is an important physiological process that eliminates cellular components through degradation in proteasomes and the lysosomes promoting cellular repair. However, its role in hepatocellular injury is not well defined [57]. Rather than receptor-mediated activation of the degradational process, autophagy occurs in response to the energy and nutrient state of a cell to maintain the basic cell functions. Misfolded proteins and damaged subcellular organelles such as mitochondria, endoplasmic reticulum, or peroxisomes can be removed insuring cell survival and tissue homeostasis. A strong stimulus to induce autophagy in hepatocytes is nutrient deprivation which results in break down of cellular proteins to provide cells with the necessary nutrients. Insulin withdrawal induces autophagy while insulin exerts hepatoprotective effects and prevents autophagy through activation of Phosphoinositide 3-kinases (PI3 K) and Akt [38].

Inflammation—the road to insulin resistance and cell death

Insulin signaling is inhibited through inhibitory phosphorylation of IRS-1 or -2 at serine residues which occurs upon activation of pro-inflammatory signaling pathways such as TNF, as discussed above. The serine kinases that mediate phosphorylation are critically involved in the development of insulin resistance and among these PKC, JNK, and IKK are also activated in response to inflammation and cell death signaling. The close relationship of metabolic processes, inflammation, and the immune system which incorporate the same signaling molecules can be traced back in evolution. In drosophila melanogaster the fat body incorporates the homologues of the mammalian liver, hematopoietic and immune system. The fat body serves as an energy sensor, ensuring nutrient availability, coordinates survival and immune responses, and functions as an innate immune system with recognition of antigens, phagocytosis of pathogens and production of antimicrobial peptides [58, 59]. This developmental heritage may underlie the highly overlapping biological repertoire of these organs, their effects on metabolic and immune cells, and the close link between immune and metabolic processes.

The dichotomy of the cellular response to TNF makes this molecule the master switch that determines, whether survival and insulin resistance, as observed in the context of inflammation, or cell death from apoptosis occurs. The following section will discuss the importance of the inflammatory mediators that are critically involved in regulating these cellular responses.

TNF and NF-κB

The biological effects of TNF are diverse and include regulation of liver injury and repair, signaling in an immune response to pathogens, and hepatocyte proliferation. On the other hand TNF is well known to promote insulin resistance and inflammation, and integrates metabolic, inflammatory and survival signaling pathways [60, 61]. The effect of TNF on hepatocytes depends on the cell’s transcriptional activity and especially NF-κB activity resulting in either activation of pro-death kinases and caspases, or factors that block apoptosis, promote tissue repair and insulin resistance (Fig. 1). In patients, elevated levels of TNF were found in acute liver failure and chronic viral, alcoholic, or non-alcoholic liver injury, as well as in obese and insulin-resistant individuals (see below) [46]. The importance of TNF in acute liver injury was shown in studies that utilized galactosamine and LPS (Gal/LPS) or carbon-tetrachloride (CCl4), both TNF-dependent models, in which neutralizing antibodies to TNF or deletion of TNF-R1 prevented liver injury [62, 63]. Additionally, mice lacking TNF-R1 exhibited reduced proliferation and liver mass following partial hepatectomy related to impaired release of IL-6 [64, 65]. In chronic viral hepatitis TNF can suppress viral replication by non-cytopathic and cytopathic mechanisms promoting clearance of virus-infected hepatocytes [66, 67]. Observations in obese rodents have stressed the importance of TNF in the pathogenesis of the metabolic syndrome and insulin resistance. In different genetic and diet-induced obesity models, increased expression of TNF was observed and neutralization of TNF or deletion of the TNF-receptor improved insulin resistance in these animals [60, 68]. Obese patients suffering from insulin resistance and non-alcoholic steatohepatitis (NASH) likewise exhibited increased levels of TNF and CD95/Fas ligand and the corresponding receptors in the liver. Additionally, increased numbers of apoptotic hepatocytes and apoptosis-associated degradation products were detected in these patients, stressing the importance of apoptotic liver injury in NASH [69, 70]. While the role of TNF and TNF-R1 in liver injury and obesity is well established, the intracellular signaling molecules that mediate TNF effects are less defined (Fig. 1). TNF-R1 mediated recruitment of TRAF-2 and RIP and subsequent JNK activation induces both, pro-injurious and insulin resistance pathways [71]. In contrast, recruitment of TRADD/FADD has so far mainly been implicated in apoptosis, but not insulin resistance [72].

The main source of TNF in liver injury, regeneration, and apoptotic cell death appears to be infiltrating and residential inflammatory cells in the liver, while in obesity the predominant source of TNF is ascribed to infiltrating inflammatory cells in white adipose tissue. In obese mouse models increased numbers of infiltrating macrophages were found in the adipose tissue, which were involved in the sustained release of cytokines [73], and in an experimental system, elevated levels of free fatty acids propagated the release of TNF from macrophages [74]. Thus while white adipose tissue was for a long time regarded as an inactive energy storage, it is now considered a highly active endocrine organ regulating energy homeostasis, insulin sensitivity and inflammatory responses by releasing cytokines and adipocytokines [75]. The release of cytokines in obesity appears to occur in response to gut-derived endotoxins, especially lipopolysaccharide (LPS) which is part of the bacterial cell wall and can be absorbed through the intestinal mucosa. The role of cytokine release from visceral white adipose tissue deserves special interest in the context of liver injury and hepatic insulin resistance since portal blood flow supplies the liver exclusively with the mediators secreted from visceral fat depots. The observation that LPS contributes to the induction of TNF in obesity came from mice which produced higher levels of inflammatory cytokines after feeding a high fat diet [76]. Augmented lymphocyte recruitment to the liver resulting in an increased inflammatory response to LPS was also observed [77]. On the other hand, decontamination of the gut with antibiotics to prevent increased LPS absorption exerted a protective effect on hepatic injury in leptin-deficient ob/ob mice [78].

The critical role of NF-κB in the regulation of cellular survival and regeneration or cell death has lead to intense research of its regulation. As described above, transcriptional activity of NF-κB is regulated through IKK-dependent degradation of IκB. The IKK complex consists of two catalytically active kinases IKK1 (IKKα) and IKK2 (IKKβ), and the regulatory kinase IKKγ (NEMO) [79]. Numerous other kinases have been implicated in NF-κB phosphorylation, including mitogen-activated protein kinases (MAPK), PKC isoforms, casein kinase II and the nuclear DNA repair enzyme poly-(ADP-ribose) polymerase-1 (PARP-1) [80–82]. However, their contribution to NF-κB activation and signaling in the liver is disputed. The principal influence of NF-κB on the regulation of obesity was substantiated by the observation that the use of anti-inflammatory drugs that inhibited IKK activity was blocking the development of insulin resistance in obese mice [83]. Genetic studies examining constitutive activation of NF-κB in the liver described a phenotype that is characterized by hyperglycemia and hepatic insulin resistance. Additionally, these animals exhibited systemic insulin resistance with decreased glucose uptake in skeletal muscle, which was reversed by inhibition of NF-κB activation [84]. The chronic activation of NF-κB in these mice resulted in increased levels of pro-inflammatory cytokines, especially TNF and Interleukin (IL)-6 [84]. Obesity-inducing diets, e.g. high fat or high carbohydrate, activated NF-κB in mice and pharmacological inhibition of NF-κB through inactivation of IKKβ reduced the degree of cellular injury to hepatocytes, involving decreased inflammation in adipose tissue and the liver [85, 86]. The specific deletion of IKKγ (NEMO) in hepatocytes resulted in chronic hepatitis with increased liver steatosis, inflammation, and apoptosis. The compensatory, reactive proliferation of hepatocytes in this context promoted dysplasia, and resulted in the development of HCC [87]. These partly opposing results are thought to result from different degrees of NF-κB inhibition in these models and the cell types affected. The sources of the pro-inflammatory cytokines in the liver are residential macrophages, the Kupffer cells—importantly, Kupffer cells responded to exposure with free fatty acids with increased expression of TNF and IL-6, and blockade of NF-κB activation in these cells prevented the secretion of proinflammatory cytokines [88]. In animal models of hepatic steatosis and inflammation, activation of Kupffer cells promoted cellular injury, thus, stressing the importance of this cell type in TNF secretion and activation of NF-κB [89]. The regulation of cytokine release from Kupffer cells is in parts mediated by natural killer T cells (NKT), members of the innate immune system, and a dysregulation of NKT cells was observed in hepatic steatosis induced by leptin deficiency in ob/ob mice or dietary induced hepatic steatosis [90, 91]. In the context of liver injury, activation of NF-κB is important to prevent cell death from TNF, through expression of anti-apoptotic proteins, such as Bcl-2 family members, FLIP, XIAP and c-IAP proteins [54, 55]. Thus, the cellular integrity of hepatocytes upon TNF exposure from inflammation is protected through activation of NF-κB, however at the price of insulin resistance.

Another prominent cytokine that regulates liver physiology and is released from white adipose tissue in obesity is IL-6 [92]. IL-6 is a member of the pro-inflammatory cytokine family that has been implicated in liver injury, liver regeneration, and hepato-carcinogenesis. Upon binding of IL-6 to its receptor glycoprotein 80 (gp80) a second membrane bound glycoprotein, gp130, is recruited which initiates the activation of Janus kinases (JAKs), leading to the activation of signal transducers and activators of transcription (STAT) proteins and MAPK pathways [93]. Hepatocyte regeneration from IL-6 is driven through activation of PI3 K and Akt, independently from insulin signaling pathways.

While the role of TNF in apoptosis and insulin signaling has been extensively studied, the importance of other members of the TNF superfamily is less known. TNF-like weak inducer of apoptosis (TWEAK) is a pro-inflammatory mediator, which inhibited insulin-induced activation of Akt involving NF-κB activation and promoted increased inhibitory serine phosphorylation of IRS-1 in hepatoma cell lines [94]. Additionally, activation of JNK occurred in response to TWEAK making it a potential candidate for further studies in patients with obesity.

Serine kinases—mediators of insulin resistance and cell death

Serine kinases are at the interface of insulin resistance and liver injury through apoptosis. As discussed above, one potential reason for this critical role is the evolutionary tight relationship of metabolic processes and immune responses. In this context the term “met-inflammation” has been used to describe the connection between the dramatic increase of inflammation-associated diseases and endemic obesity [95]. Among these, JNK activation was recognized to occur in the liver, muscle, and adipose tissue of obese mice and organ-specific inactivation of JNK prevented the development of diet-induced obesity and injury [96, 97]. JNK activation occurs from phosphorylation in response to a variety of cellular stressors including oxidative and endoplasmatic reticulum (ER) stress, inflammatory cytokines, especially TNF, but also metabolic stress including obesity and increased supply of free fatty acids [98]. A total of three JNK isoforms exist, however in hepatocytes only JNK1 and 2 are expressed [99]. The specific inhibition of JNK1 or JNK2 in mice has shown that these isoforms mediate differential effects in the context of hepatic steatosis, cytokine-induced apoptosis, and oxidative stress. In NASH JNK1 promoted the development of inflammation, steatosis and liver injury [97]. Activation of JNK1 in adipocytes was shown to promote hepatic insulin resistance through a mechanism of increased secretion of IL-6 from adipose tissue resulting in SOCS-3 expression in the liver, and deletion of JNK1 in adipocytes prevented hepatic insulin resistance [98]. Singh and colleagues explored the differential effects of JNK1 and JNK2 inhibition in hepatocytes: While selective knockdown of JNK1 in hepatocytes protected from the development of steatosis and insulin resistance in response to an obesity-inducing diet, deletion of JNK2 improved only insulin resistance but did not affect hepatic triglyceride accumulation [100]. Besides the regulation of insulin sensitivity, JNK is a critical regulator of acute hepatocellular injury, liver regeneration, and carcinogenesis [101]. Oxidative stress in hepatocytes causes cell death in which the MAPK Erk and the transcription factor cJun, a target of JNK, play an opposing role: the MAPK Erk protects hepatocytes, while c-Jun induces cell death of hepatocytes. Non-cytopathic concentrations of the superoxide generator menadione activate both JNK and the MAPK Erk to achieve a balance that promotes hepatocyte survival. However, high concentrations of menadione or inhibition of Erk led to a prolonged activation of JNK and c-Jun, and cell death occurred [102]. In acute liver injury from Gal/LPS, the lack of JNK2 in hepatocytes prevented cell death and liver injury, while deletion of JNK1 had no effect on hepatocellular injury. Additionally, the protection from Gal/LPS mediated injury was independent of the phosphorylation of its transcription factor c-Jun but rather accompanied by reduced activation of effector caspases, implying an important regulatory function of JNK2 in apoptosis independent from its downstream transcription factor [103]. In summary, the serine kinase JNK connects injury and insulin resistance in hepatocytes in response to TNF and as such could be an important therapeutic target.

SOCS proteins act as negative regulators of inflammation counterbalancing the effects of cytokines and growth factors on inflammatory cells in response to cellular injury, and thus prevent activation of the immune system. Deletion of SOCS-1 in mice resulted in a lethal hepatitis and steatosis from aberrant T-cell activation [104]. As described above, this group of proteins induced insulin resistance independent of IRS serine phosphorylation involving the ubiquitin-dependent degradation of IRS molecules [105] and deletion of SOCS in obese mice improved insulin resistance and hepatic steatosis [106], while SOCS-3 expression promoted insulin resistance in adipocytes [107].

Adipocytokines

Adipocytokines constitute a heterogeneous group of bioactive mediators that are secreted predominantly from adipocytes in response to metabolic and inflammatory stimuli, making white adipose tissue an important endocrine organ. Prominent among these are adiponectin, leptin and resistin. Specific receptors for the different adipocytokines can be found in many tissues and the above mentioned proteins mediate differential effects, with profound impact on hepatic injury and repair, and insulin resistance, making them central mediators of apoptosis and diabetes in the liver.

Adiponectin is a proteo-hormone, secreted predominantly from adipocytes [108]. In the peripheral blood adiponectin circulates as multimers which exhibit different receptor affinity and biological function depending on molecular size [109, 110]. The predominant function of adiponectin is the regulation of glucose homeostasis through its effect on hepatic gluconeogenesis, skeletal muscle glucose uptake, and energy expenditure [111, 112]. In the central nervous system adiponectin regulates food uptake influencing obesity and insulin sensitivity [113]. In mouse models of obesity and insulin resistance, adiponectin improved hyperglycemia and hyperlipidemia, and deletion of adiponectin in mice resulted in insulin resistance [114, 115]. In line with these observations, an inverse correlation of serum adiponectin levels and the body mass index (BMI) was found in obese humans and hypo-adiponectinemia was independently associated with the metabolic syndrome and diabetes [116]. Also genetic analyses in humans showed that single nucleotide polymorphisms (SNP) of adiponectin were associated with insulin resistance and cardiovascular disease [117, 118]. Adiponectin acts through the adipo-R1 receptor with ubiquous expression and adipo-R2 receptor, which can be found in the liver [119]. Adipo-R1 expression was shown to stimulate activation of 5′AMP-activated protein kinase (AMPK) suppressing hepatic gluconeogenesis while Adipo-R2 regulates glucose uptake and glucose utilization through peroxisome proliferator-activated receptor-α (PPAR-α) [120]. Adiponectin increased the β-oxidation of free fatty acids and decreased de novo fatty acid synthesis through regulation of the rate controlling enzymes acetyl-CoA-carboxylase (ACC), AMPK and the transcription factor sterol regulatory element-binding protein (SREBP)-1c and prevented hepatic steatosis [121]. Besides the effect on glucose homeostasis, adiponectin exerts strong anti-inflammatory and anti-fibrotic effects in the liver. In patients with NASH, hypoadiponectinemia correlated positively with beta-cell dysfunction, the degree of liver fibrosis [122] and expression of C-reactive protein (CRP) in endothelial cells [123]. The mechanism by which adiponectin exerts anti-inflammatory effects involves a counter-regulatory effect on TNF production. In line with this, increased TNF serum levels were observed in adiponectin knock out mice [114] and selective deletion of adiponectin in myeloid cells increased their TNF secretion in response to LPS [124]. Therefore it was not surprising to find that adiponectin protected hepatocytes from injury and apoptosis in NASH, which is characterized by TNF-induced steatosis and insulin resistance [125]. In a profibrogenic, injurious mouse model employing chronic CCl4 adiponectin exhibited a cytoprotective effect on hepatocytes [124]. The anti-inflammatory effects of adiponectin included suppression of TNF-induced expression of vascular cell adhesion molecule-1 (VCAM-1), endothelial-leukocyte adhesion molecule-1 (ELAM-1), E-selectin, and other intercellular cell adhesion molecules [126]. Central to the counter regulatory effects of TNF through adiponectin was suppressed activation of NF-κB binding activity in different cell types and increased release of anti-inflammatory cytokines such as IL-10 [123, 127]. In summary adiponectin acts as an anti-inflammatory mediator counterbalancing TNF- effects and prevents hepatocellular injury and fibrosis (Fig 2).

Fig. 2
figure 2

Regulation of insulin resistance and apoptosis through adipocytokines. Adipose tissue secrets hormones which are critically involved in the regulation of insulin resistance and cell death in hepatocytes. Levels of adiponectin correlate inversely with the adipose tissue mass. High levels of adiponectin are detected in lean individuals, while low levels of adiponectin occur in obesity. Adiponectin counter regulates the TNF-induced pro-inflammatory effects and thus prevents insulin resistance and apoptosis in hepatocytes. Leptin acts primarily in the central nervous system and suppresses or stimulates appetite depending on the bodies energy depots and adipose tissue mass. In obesity, leptin resistance (⊗) develops, which leads to a positive energy balance with expanding adipose tissue stores. The physiological role of leptin includes insulin sensitizing and anti-inflammatory effects. In the liver, leptin activates and protects hepatic stellate cells (HSC) which are involved in chronic injury and liver fibrosis. The adipocytokine resistin promotes insulin resistance through decreasing activation of IRS signaling molecules in response to insulin. Additionally, resistin appears to be involved in mediating the inflammatory response in the adipose tissue observed in obesity, involving increased expression of chemoattractants. Adipose tissue inflammation and hyper-responsiveness of macrophages in response to TNF appears to be a main source of pro-inflammatory cytokines in obesity, promoting insulin resistance and on the other side, hepatocellular injury and apoptosis

Full size image

The polypeptide hormone leptin is encoded by the ob gene, that was first cloned in 1994 and the corresponding hormone discovered 1 year later [128]. The profound impact of leptin on energy homeostasis and maintenance of adipose tissue mass has lead to a high scientific interest [129]. High serum levels of leptin were detected in states of starvation in which a positive energy balance had to be achieved in order to replenish depleted adipose tissue stores. Today overnutrition rather than starvation challenges the human species in the western hemisphere: Failure to reduce caloric intake and lack of physical exercise in obese patients, is associated with partial or complete leptin resistance [130, 131]. Leptin regulates energy homeostasis and maintenance through its effects on the central nervous system, pancreas, liver, adipose tissue, and immune system [132]. Leptin deficiency in ob/ob mice resulted in metabolic alterations including insulin resistance, immune and inflammatory abnormalities with increased levels of pro-inflammatory TNF and imbalance of regulatory inflammatory cells, which was in part reversed through transplantation of white adipose tissue [133, 134]. At the molecular level leptin acts through binding to ob receptors either in a membrane-bound or soluble form. While the soluble receptor–ligand complex delayed clearance and thus prolonged leptin effects, membrane bound leptin did not activate the receptor [135]. The intracellular signaling events involve activation of the JAK/STAT pathway, activation of AMPK and reduced transcriptional activity of SREBP-1. Activation of JAK/STAT, AMPK, and suppression of the transcriptional activity of SREBP-1 improved insulin sensitivity through decreasing triglyceride accumulation in the liver and skeletal muscle [136–138]. Due to its immune-modulating and metabolic effects, the importance of leptin in liver injury, hepatocytes apoptosis, and fibrogenesis were thoroughly examined. Leptin decreased and prevented hepatic triglyceride accumulation through its inhibitory effects on pro-lipogenic pathways. This was of importance in the context of death receptor-mediated apoptosis, as hepatic steatosis is known to increase the susceptibility towards apoptosis-inducing ligands [89, 139]. The role of leptin in fibrogenesis is more controversial. On one hand it was shown to promote fibrogenesis through transforming growth factor β (TGFβ) in sinusoidal endothelial and Kupffer cells in response to chronic liver injury from CCl4 [140] and prevented apoptosis in HSC through activation of extracellular-regulated kinase (Erk) and Akt [141]. Additionally, leptin increased the expression of tissue inhibitor of metalloproteinase 1 (TIMP-1) via the JAK/STAT pathway in HSCs to promote fibrogenesis [142] and high levels of leptin were detected in liver cirrhosis [143]. On the other hand, in patients with NASH no correlation of leptin and liver fibrosis or inflammation could be observed [144]. Thus, while leptin is capable of decreasing food uptake and preventing hepatic steatosis, in obesity leptin resistance develops and the protective effects of leptin on metabolic processes are lost, while injurious effects on hepatocytes contribute to hepatocellular damage (Fig. 2).

Resistin is a 12-kDa protein that is expressed and secreted from adipose tissue. In skeletal muscle of mice it promoted insulin resistance, hyperglycemia and hyperlipdemia [145, 146], involving regulation of key enzymes of hepatic glycogen synthesis [147] and decreased activation of IRS-1 in response to insulin [148, 149]. Likewise, mice that developed obesity-associated NASH exhibited increased serum levels of resistin [150]. However, the role of resistin in humans is not fully understood yet: In patients with NASH a positive correlation of resistin with the degree of inflammation in the liver was found [151] and in acute alcoholic liver injury resistin was expressed at high levels in the liver [152]. The injurious effects of resistin on hepatocytes appeared to be predominantly from activation of HSC, which in response to resistin showed activation of NF-κB, expression of monocyte chemoattractant protein-1 (MCP-1) and the pro-inflammatory cytokine IL-8 [152].

Summary

The incidence of insulin resistance is increasing parallel to the endemic occurrence of obesity. Insulin resistance causes a state of chronic, low-grade inflammation driven by the activation of inflammatory cells in adipose and other tissues, including the liver. The key players of insulin resistance include TNF, NF-κB, JNK, and adipocytokines which are also involved in hepatocellular injury from apoptosis. The connection between insulin resistance and end organ injury due to inflammation and activation of cell death signaling pathways can be traced back in evolution to a point where regulation of metabolic processes and a host response to pathogens was represented within the same organ system. The redundancy of the signaling pathways will make it unlikely that targeting a single molecule will provide cure from insulin resistance and the associated liver disease. In contrast, understanding the complex nature of the interactions will provide the basis for identifying therapeutic approaches which will most likely include molecular interventions in both insulin resistance and cell death signaling pathways.