Obesity is a new global pandemic, with growing incidence and prevalence. This disease is associated with increased risk of several pathologies, including diabetes, cardiovascular diseases, and cancer. The mechanisms underlying obesity-associated metabolic changes are the focus of efforts to identify new therapies. Stress-activated protein kinases (SAPK), including cJun N-terminal kinases (JNKs) and p38, are required for cellular responses to metabolic stress and therefore might contribute to the pathogenesis of obesity. Tissue-specific knockout models support a cell-type-specific role for JNK isoforms, in particular JNK1, highlighting its importance in cell homeostasis and organ crosstalk. However, more efforts are needed to elucidate the specific roles of other JNK isoforms and p38 family members in metabolism and obesity. This review provides an overview of the role of SAPKs in the regulation of metabolism.
The totality of biochemical reactions that a cell requires to grow, survive and respond to external stimuli is known as metabolism. These processes are classified into two interdependent groups. Catabolism is the set of reactions that break down complex molecules to produce energy. This process produces NADH or ATP, which are required to synthesize macromolecules such as proteins, lipids and nucleic acids. This set of synthetic reactions is known as anabolism. Anabolism and catabolism play opposing roles in the maintenance of cell functions, but these processes are tightly coordinated in order to preserve cell homeostasis. Endocrine hormones, such as insulin, finely regulate the molecular pathways involved in the activation of anabolic and catabolic enzymes. Whenever this coordinated regulation is lost, anabolism and catabolism become unbalanced, giving rise to metabolic disorders. Obesity, the most common metabolic disorder worldwide, develops when energy intake exceeds energy expenditure (Speakman 2010). Obesity is a major cause of morbidity and mortality, and is associated with increased risk of type 2 diabetes mellitus, heart disease, metabolic syndrome, hypertension, stroke, and certain forms of cancer (Haslam & James 2005). The design of new treatments requires knowledge of the diverse molecular mechanisms that contribute to obesity and its associated diseases.
MAPK signalling cascades transduce a variety of extracellular signals that regulate cellular responses (Sabio & Davis 2014). Three groups of MAPKs have been identified: the ERK, the p38 MAPKs, and the cJun N-terminal kinases (JNK). ERKs are mainly activated by mitogens and differentiation signals, JNK and p38 MAPK are activated by stress stimuli and are collectively named stress-activated protein kinases (SAPK) (Fig. 1) (Sabio & Davis 2014). The SAPK family is composed of the four isoforms of p38 (p38 alpha, beta, gamma and delta) and the three isoforms of JNK (JNK1, 2, and 3) (Paul et al. 1997). While the JNK pathway has been extensively studied and JNK1 has been implicated in the development of obesity and insulin resistance (Sabio & Davis 2010), the role of p38 MAPKs in this context has received less attention. Obesity results in a low-grade inflammation state (Neels & Olefsky 2006), with increased levels of proinflammatory cytokines such as tumour necrosis factor alpha (TNFα) and interleukin 6 (IL6) and of circulating free fatty acids (FFA), causing the activation of stress-induced mechanisms in the cell (Hotamisligil 2006).
cJun N-terminal kinases
Three genes encode the members of JNK family. While Jnk1 and Jnk2 are ubiquitously expressed, Jnk3 is specifically expressed in brain, testis and heart (Davis 2000, Chang & Karin 2001). Each JNK member can be expressed in several splice variants, which can be separated in two groups based on their size: the short forms of about 46 kDa (JNK1α1, JNK1β1, JNK2α1 JNK2β1 JNK3α1) and the long forms of about 54 KDa (JNK1α2, JNK1β2, JNK2α2 JNK2β2 JNK3α2) (Gupta et al. 1996, Yu et al. 2004). Many cells types express at least four JNK splice variants, with JNK1 variants predominating over JNK2 variants among the short forms and the reverse pattern observed for the long forms (Gupta et al. 1996, Yu et al. 2004). The function and the exact modulation of each of these splice variants is still unsolved.
JNKs are activated in response to a wide range of stimuli, such as cytokines (IL6, TNFα), environmental stresses (hypoxia, u.v. and ionizing radiation), toxins, drugs and metabolic changes, including obesity and hyperlipidaemia. JNKs are directly activated by two upstream MAP kinase kinase (MAP2K) enzymes, MKK4 and MKK7, through a dual phosphorylation on tyrosine and threonine in the conserved Thr-Pro-Tyr motif in their activation loop (Paul et al. 1997). The more than 50 proteins identified as JNK substrates include c-Jun, which once phosphorylated can dimerize with JunB, JunD or c-Fos to form the transcription factor activator protein-1 (AP-1) (Hotamisligil et al. 1996a); insulin receptor substrate 1 (IRS1) (Kim et al. 2004b); c-myc; p53, and numerous transcription factors (Davis 2000, Chang & Karin 2001, Bogoyevitch & Kobe 2006). Through the phosphorylation of their targets, JNKs induce a specific cell response to the specific stimulus, characterized by apoptosis, cell proliferation, and cell migration.
Despite its importance for brain function (Pirianov et al. 2007, Yoon et al. 2012), the JNK3 isoform as yet has no demonstrated role in metabolism. In contrast, a large body of literature attests to the key roles of JNK1 and JNK2 isoforms as regulators during obesity-induced inflammation and metabolic disorders (Tuncman et al. 2006, Singh et al. 2009, Czaja 2010, Han et al. 2013). The JNK pathway is strongly activated in liver, adipose tissue and muscle during dietary and genetically-induced obesity (Hirosumi et al. 2002). Hirosumi et al. (2002) showed that mice with whole-body knockout of JNK1, but not JNK2, are protected against obesity and insulin. This finding correlates with a reduced total JNK activation in JNK1 knockout, whereas JNK2 knockout mice do not show this effect. Moreover, JNK1 knockout mice gain less weight (Hirosumi et al. 2002) and are resistant to liver steatosis induced by a diet deficient in methionine and choline (Schattenberg et al. 2006). Here, we provide an overview of the roles of JNK1 and JNK2 in metabolism and metabolic disorders (Fig. 2).
In mammals, the liver plays an essential role in metabolism, and is central to the maintenance of glucose and lipid homeostasis in the body. Hepatic insulin resistance and non-alcoholic fatty liver disease (NAFLD) are the most frequent liver consequences of the obese state. Using adenoviruses to overexpress or downregulate JNK, Nakatani et al. (2004) defined a central role for the hepatic JNK pathway in determining whole-body insulin sensitivity. These authors showed that hepatic overexpression of WT JNK induces insulin resistance in non-diabetic mice, whereas suppression of the JNK pathway in liver ameliorated insulin resistance in genetically obese mice (Nakatani et al. 2004). Moreover, studies in whole-body JNK1-knockout mice have shown that this kinase controls lipid accumulation and lipid uptake in liver (Imbernon et al. 2013). These findings indicate that the role of JNKs isoforms can differ between cell types. In fact, the phenotype of whole-body JNK1-knockout mice (Hirosumi et al. 2002) and adenovirus-treated mice (Nakatani et al. 2004) could be due to the ablation of JNK1 in parenchymal cells or extra hepatic tissues, highlighting the importance of cell and organ crosstalk. Hepatocyte-specific deletion of JNK1 provided further information about its role in liver steatosis and insulin resistance during obesity. Intriguingly, these mice displayed increased insulin resistance, glucose intolerance and hepatic steatosis, showing that JNK1 plays a protective role in hepatocytes (Sabio et al. 2009). In contrast, hepatocyte-specific simultaneous knockout of JNK1 and JNK2 induces systemic protection against high fat diet (HFD)-induced insulin resistance (Vernia et al. 2014). Vernia et al. demonstrated that JNK could indirectly inhibit the activity of the nuclear peroxisome proliferator-activated receptor alpha (PPARα), thus preventing production of fibroblast growth factor 21 (FGF21). FGF21 plays a pivotal role in control ketogenesis, insulin sensitivity, glycaemia and obesity. In consequence, they found that lack of JNK1/2 in hepatocytes protects against obesity-induced diabetes (Vernia et al. 2014). JNK1 modulates liver insulin resistance and steatosis through two distinct molecular actions. On one hand, JNK1, upon activation by proinflammatory cytokines (e.g. TNFα) and circulating FFAs, phosphorylates the IRS1 on serine 307, preventing its interaction with the insulin receptor. This blocks insulin cellular signalling, inducing insulin resistance (Hotamisligil et al. 1996b, Aguirre et al. 2000). On the other hand, JNK1 is necessary in hepatocytes for the homeostatic maintenance of gluconeogenesis and lipogenesis and the prevention of hepatic insulin clearance due to the elevated expression of insulin receptor in this cell type (Sabio et al. 2009).
The most important organ for lipid storage is adipose tissue, the only tissue with high capacity to expand in a non-transformed state. However, adipocytes are not merely fat storage cells, in fact play a central role in metabolism and organismal homeostasis through the secretion of adipokines (Gavrilova et al. 2000). In the obese state, adipocyte enlargement induces molecular and cellular alterations that affect whole-body metabolic homeostasis. Increased lipolysis in adipocytes leads to increased hyperlipidaemia and lipid accumulation in peripheral organs, in particular muscle and liver. Moreover, adipocytes begin to produce a number of proinflammatory cytokines, such as IL6, TNFα, monocyte-chemoattractant protein-1 (MCP-1), inducible isoform nitric oxide synthase (iNOS), transforming growth factor-beta1 (TGF-β1), and plasminogen activator inhibitor-1 (PAI-1) (Hotamisligil et al. 1993). Obese adipose tissue also recruits macrophages that contribute to the generation of a chronic inflammation state and insulin resistance (Weisberg et al. 2003, Xu et al. 2003). Adipocytes are thus target of stress stimuli during obesity, and, as mentioned previously, genetically- and dietary-induced obesity both activate JNK in adipose tissue (Hirosumi et al. 2002). The scaffold protein JNK interacting protein 1 (JIP1) has been shown to play a central role in JNK activation in adipose tissue (Jaeschke et al. 2004). JIP1−/− mice are resistant to diet-induced obesity and show less severe insulin resistance. JNK cannot be activated in the adipose tissue and muscle of JIP1−/− mice, preventing the acquisition of insulin resistance in those tissues. However, the livers of JIP1−/− mice have no defect in JNK activation, which may account for the incomplete protection against insulin resistance in these animals. This demonstrates that adipose tissue protection from insulin resistance is cell specific and it is not due to a whole body protection from obesity (Jaeschke et al. 2004). Mice with adipocyte-specific JNK1-deficiency have been used to analyse the role of JNK1 in adipose tissue (Sabio et al. 2008). When fed a high-fat diet, these mice show a similar body-mass gain to their control counterparts. However, the adipocyte JNK1-knockout mice are protected against insulin resistance in adipose tissue. This phenotype could be due to the absence of JNK1-mediated inhibitory phosphorylation of IRS1. Surprisingly, knockout of JNK1 in adipose tissue also induces protection against insulin resistance in liver (Sabio et al. 2008). Adipose JNK1-knockout mice exhibit a strong reduction in adipose tissue production of IL6, reflected in decreased levels of circulating IL6. This proinflammatory cytokine is known to induce hepatic insulin resistance (Klover et al. 2003, Kim et al. 2004a). In conclusion, JNK1 in adipose tissue regulates local insulin sensitivity through IRS1 phosphorylation and also controls adipocyte-IL6 production and adipocyte crosstalk with other organs. In the obese state, adipose JNK1 induces hepatic insulin resistance through IL6 production (Sabio et al. 2008).
Obesity induces a low-grade chronic inflammation that could be partly responsible for the observed metabolic changes (Hotamisligil 2006, Neels & Olefsky 2006). Increasing food intake increases adipose tissue fat storage, which induces adipocyte enlargement, hypoxia, stress, cell death, and proinflammatory cytokine production. These changes have been proposed to induce immune-cell recruitment to adipose tissue and produce a chronic activation of the immune response (Ferrante 2013). However, how fat deposition in adipose tissue activates the immune system is not completely understood. The obese state also induces immune activation in organ-resident macrophages, in particular Kupffer cells in the liver, probably inducing insulin resistance (Lanthier et al. 2010). In order to assess the role of JNK in the immune system in obesity and obesity-induced disorders, several studies have examined mice lacking JNK proteins in myeloid or bone-marrow-derived specific knock out mice. Zhang et al. (2011) examined the effect of a high-fat diet on transgenic mice overexpressing dominant-negative JNK in adipose tissue and macrophages. These mice showed reduced weight gain, insulin resistance, glucose intolerance and hepatic steatosis. Moreover, these mice had smaller-than-normal adipocytes, reduced macrophage infiltration in adipose tissue, and less-severe whole-body inflammation (Zhang et al. 2011). The authors concluded that JNK isoforms are responsible for the crosstalk between adipose tissue and macrophage infiltration, and showed that abrogation of their expression in both cell types is necessary to obtain the same phenotype seen in JNK1 full body knock out mice (Zhang et al. 2011). However, these results do not fully elucidate the role of JNK proteins expressed in hematopoietic cells. Solinas et al. (2007) used adoptive transfer to demonstrate that JNK1 deletion in non-hematopoietic tissues protects against weight gain and, partly as a consequence, from insulin resistance. Moreover, they showed that hematopoietic JNK1 removal reduces obesity-induced inflammation and protects against HFD-induced insulin resistance without affecting obesity (Solinas et al. 2007). However, another group reported that JNK1-deficient bone marrow transplantation was not sufficient to induce protection against insulin resistance and macrophage infiltration of adipose tissue in a WT recipient, whereas this protection was induced by the absence of JNK1 in parenchymal cells (Vallerie et al. 2008). Taken together, these results highlight the importance of understanding the role of JNK2 in the myeloid lineage. To address this question, the Davis laboratory generated a myeloid-cell-specific knockout mouse model for JNK1 and JNK2 (Han et al. 2013). When fed a HFD, these mice show the same obesity and hyperlipidaemia responses as WT controls, but are more insulin sensitive (Han et al. 2013). The myeloid JNK1/2-knockout mice also show less macrophage infiltration and decreased M1 polarization in adipose tissue. M1 polarization was also decreased in hepatic macrophages, accompanied by increased M2 polarization, highlighting an important role for JNK in M1 polarization. In vitro experiments further confirmed an intrinsic defect in M1 polarization (Han et al. 2013). Finally, a role has been described for JNK2 in T cells in type 1 diabetes (Jaeschke et al. 2005). Non-obese diabetic (NOD) mice lacking JNK2 exhibit a decrease in spontaneous diabetes and insulitis, probably due to increased Th2 polarization of CD4+T cells (Jaeschke et al. 2005). The authors of this study reported that this T-cell polarization effect was not seen in JNK1−/− NOD mice, indicating a specific role for JNK2 in the regulation of T-cell polarization (Jaeschke et al. 2005).
Skeletal muscle is a very large organ that requires abundant available energy to carry out its functions. Myocytes therefore play a major role in insulin-induced glucose disposal and are sensitive to obesity-induced insulin resistance. Evidence suggests that circulating FFAs, whose levels are increased in obesity, might be responsible for skeletal muscle insulin resistance (Shulman 2000, Roden 2004). FFAs induce the phosphorylation cascade that activates JNKs, suggesting a possible a role for JNKs in skeletal muscle insulin resistance. Moreover, JNK1 is activated in muscle from HFD-fed mice (Hirosumi et al. 2002). Muscle-specific JNK1-knockout mice reveal a role for this isoform in the induction of insulin resistance in skeletal muscle fed a HFD (Sabio et al. 2010a). These mice are sensitive to HFD-induced obesity. However, they are protected against obesity-induced insulin resistance and show increased glucose uptake by muscle, probably as result of a reduction in IRS1 serine phosphorylation. Surprisingly, adipose tissue and liver are not positively affected by JNK1 deficiency in skeletal muscle (Sabio et al. 2010a). Instead, these mice show increased hepatic steatosis and circulating levels of triglycerides, due to a reduction in lipoprotein lipase (Lpl) gene expression in muscle. Moreover, the augmented blood levels of triglycerides induce an increase in adipose tissue macrophage infiltration (Sabio et al. 2010a). Interestingly, inhibiting JNK in muscle by overexpressing heat shock protein 72 (HSP72) improves HFD-induced hyperglycaemia and hyperlipidaemia (Chung et al. 2008). Moreover, it prevents glucose intolerance and insulin resistance, and reduces HFD-induced weight gain (Chung et al. 2008). Taken together, these results indicate that lack of JNK1 expression in muscle modulates insulin sensitivity in a cell-autonomous manner but does not influence insulin sensitivity in other tissues (Sabio et al. 2010a). However, inhibiting total JNK activity in muscle can ameliorate the whole-body response to HFD (Chung et al. 2008). It would be interesting to study the effect of JNK2 deficiency in muscle to identify its possible role in tissue crosstalk.
Pancreatic β cells
Insulin is produced by β cells in the pancreatic islets of Langerhans. In a healthy state, β cells, which make up 65–80% of the cells in the islet, respond to increases in serum glucose by secreting insulin, which induces glucose uptake in peripheral organs. Chronic hyperglycaemia induces changes in β-cell gene expression that lead to cell hypertrophy and loss of differentiation (Jonas et al. 1999). Glucose-induced insulin gene expression is also inhibited by saturated fatty acids (which are increased in obesity) through the activation of JNK in β cells, resulting in the phosphorylation and inhibition of IRS1 and IRS2 (Solinas et al. 2006). Moreover, overexpression experiments have shown that activation of JNK by oxidative stress leads to β-cell failure (Kaneto et al. 2002, 2005). JNK-interacting protein 1, a scaffold protein that regulates JNK activity (Whitmarsh et al. 2001), has been identified as a potential regulator of type 2 diabetes in humans (Waeber et al. 2000) and has been shown to mediate β-cell apoptosis (Haefliger et al. 2003). Despite these results, experiments in animals lacking JIP1/JIP2 protein indicate that these proteins are not essential for β-cell viability (Standen et al. 2009). Activation of JNK in β cells by overexpression of constitutively active MKK7 results in impaired insulin production (Lanuza-Masdeu et al. 2013), a result consistent with previous work showing that JNK2−/− β cells in non-obese mice are protected against apoptosis induced by T cells in a model of autoimmune type 1 diabetes (Jaeschke et al. 2005). All these results indicate a specific role of JNK in β cells as a regulator of insulin production and secretion, and prolonged JNK activation, as in the obese state, can induce islet dysfunction.
In the CNS, the onset of hypothalamic resistance to leptin and insulin is a key event in the development of obesity and impaired glucose homeostasis and systemic immune responses (Belgardt & Bruning 2010, Tschop et al. 2010, Vogt & Bruning 2013). Both hormones act on hypothalamic neurons to decrease food intake and hepatic glucose production and to increase energy expenditure (Obici et al. 2002, Belgardt & Bruning 2010). Two neuronal populations, each characterized by the expression of specific neuropeptides, have potent effects on energy homeostasis (Contreras et al. 2015, Lopez & Tena-Sempere 2015). One population consists of the POMC and CART neurons, which provide a strong anorexigenic effect, decreasing food intake and body weight. In contrast, the second population of AgRP and NPY neurons has a potent orexigenic effect, increasing food intake. Since the first evidence demonstrating that brain-specific deletion of insulin receptor (InsR) was associated with diet-sensitive obesity (Bruning et al. 2000), many reports have tried to identify the neurons and mechanism by which this hormone affects the brain. In this context, the failure of conditional depletion of JNK1 in several peripheral insulin-sensitive tissues to recapitulate the phenotype of whole-body JNK1 knockout mice (Sabio et al. 2008, 2009, 2010a,b) suggests the possibility that JNK1 activation in the CNS might contribute to the maintenance of energy balance and glucose metabolism. JNK activation in hypothalamus during obesity has been linked to endoplasmic reticulum stress, inflammation, and hyperlipidaemia (Ozcan et al. 2004, Prada et al. 2005). In 2010, two groups independently generated two brain-specific conditional JNK1 knockout models (Belgardt et al. 2010, Sabio et al. 2010b) using the Nestin-cre system, which is used to ablate a gene of interest in neurons and astrocytes in the CNS (Tronche et al. 1999). Both studies showed that these mice are protected against HFD-induced obesity, insulin resistance and glucose intolerance at the central and peripheral level (Belgardt et al. 2010, Sabio et al. 2010b). Reduced body-weight gain correlated with increased energy expenditure and locomotor activity and a higher body temperature (Belgardt et al. 2010, Sabio et al. 2010b). Moreover, the mice had depressed lipid accumulation in brown fat and elevated plasma levels of T3 and T4 thyroid hormones, which correlated with high levels of thyrotropin-releasing hormone (TRH) in hypothalamus. These data indicate that JNK1 in CNS controls body homeostasis by regulating the hypothalamic–pituitary–adrenal axis (Belgardt et al. 2010, Sabio et al. 2010b). Two newly generated mouse models are helping to further clarify the role of JNK1 in the CNS: deletion of JNK1 and JNK2 in pituitary gland and conditional overexpression of constitutively active JNK in AgRP neurons (Vernia et al. 2013, Tsaousidou et al. 2014). Vernia et al. found that lack of JNK1 and 2 in the anterior pituitary gland protects against weight gain during obesity and diabetes. Moreover, these mice have elevated energy expenditure and circulating levels of T3, T4 and thyroid stimulating hormone (TSH). The study shows that JNK1/2 regulate expression of Dio2, which encodes the enzyme responsible for T4 to T3 conversion in the pituitary gland. Impairment of this conversion reduces the negative feedback regulation of the expression of TSH, resulting in sustained production of TSH, T3 and T4 (Vernia et al. 2013). In the second model, mice expressing constitutively active JNK1 in AgRP neurons become obese when fed a HFD and are leptin resistant at the neuronal and systemic level (Tsaousidou et al. 2014). These results indicate that JNKs, in particular JNK1, play a pivotal role in central nervous control of obesity. Further efforts are needed to address the role of JNK3, a mainly CNS isoform, in metabolism and energy homeostasis.
The four p38 MAPK isoforms – α, β, γ, δ – are encoded by different genes and have different tissue expression patterns (Sabio & Davis 2014). The p38α isoform is widely expressed, but is less abundant in the brain, where p38β MAPK is the major isoform (Beardmore et al. 2005). The p38δ isoform is highly expressed in a limited number of tissues, including neutrophils and endocrine glands (Sumara et al. 2009, Cuadrado & Nebreda 2010, Ittner et al. 2012, Gonzalez-Teran et al. 2013), while p38γ kinase is expressed in all tissues, with high levels found in muscle (Beardmore et al. 2005).
All p38 isoforms are activated, in response to appropriate stimuli, by dual phosphorylation in the activation loop sequence Thr-Gly-Tyr. p38 MAPKs can be activated by MKK3, MKK4 or MKK6 in cell-free systems, but their activation in intact cells is primarily mediated by MKK3 and MKK6 (Enslen et al. 2000, Brancho et al. 2003, Remy et al. 2010). However, most of these studies were performed in immortalized fibroblasts and how specific p38 MAPK isoforms are activated in response to different stimuli in other cells types is largely unknown. The function of p38α has been studied in several physiological processes, and although no specific role for this kinase has been defined in metabolism, p38α is known to phosphorylate several key proteins involved in glucose and lipid metabolism (Fig. 3).
Although the specific role of the different p38 isoforms in liver metabolism has not been studied, several p38 MAPK substrates have been implicated in metabolic processes. Here we summarize these substrates and current knowledge about the potential action of p38 isoforms in liver metabolism. Downstream targets of p38 include two important transcription factors involved in gluconeogenesis: transcriptional PPAR gamma coactivator-1 alpha (PGC1α) and CREB. Upon phosphorylation, PGC1α and CREB activities are enhanced, and they bind to sites in the promoter regions of the genes encoding phosphoenolpyruvate carboxykinase kinase (PEPCK) and glucose-6 phosphatase, activating transcription of these gluconeogenic genes (Deak et al. 1998, Puigserver et al. 2001, Cao et al. 2005). Experiments with pharmacological inhibitors and p38-targeting shRNA have suggested that p38 might modulate gluconeogenesis via this mechanism (Cao et al. 2005). PGC1α also co-activates many other transcription factors, including PPARα and FOXO1. Furthermore, p38 has been proposed to directly phosphorylate PPARα, enhancing its activity (Juge-Aubry et al. 1999). p38 also phosphorylates CCAAT-enhancer-binding protein alpha (C/EBPα) (Qiao et al. 2006), a transcription factor that regulates glucose homeostasis (Yang et al. 2005). Phosphorylation on serine enhances C/EBPα transactivation activity and increases PEPCK gene expression (Qiao et al. 2006). A more recent study showed that p38 phosphorylates the spliced form of X-box binding protein 1 (Xbp1s), promoting its nuclear translocation, reducing endoplasmic reticulum stress, and improving glycaemia during obesity (Lee et al. 2011). All these studies point to a potentially important role of p38 kinases as regulators of glucose homeostasis in the liver. However, extensive research in conditional knockout and transgenic mouse models will be needed to define the function of each p38 isoform in liver homeostasis.
Adipose tissue formation depends on tight regulation of cell growth, proliferation and differentiation (Gehart et al. 2010). It has been proposed that p38 might modulate adipocyte differentiation through the phosphorylation of C/EBPβ (Engelman et al. 1998), a key transcription factor for adipose tissue differentiation (Tanaka et al. 1997). In vitro experiments with p38α/β inhibitors have shown that p38 phosphorylates and activates C/EBPβ, enhancing PPARγ transcription and promoting adipocyte differentiation. It also has been postulated that p38α regulates brown adipose tissue differentiation by controlling uncoupling protein 1 (UCP1) expression levels through the phosphorylation of the transcription factors ATF2 and PGC1α (Cao et al. 2004, Tseng et al. 2008). However, more experimental work is needed in knockout and conditionally deficient animals in order to define the role of p38 MAPK in adipose tissue in vivo.
Pancreatic β cells
Although it has been speculated that p38α might promote β cell failure, a part from in vitro inhibitor studies, there are only a few supporting evidences (Gehart et al. 2010). A very elegant recent study showed that p38δ controls insulin secretion through the phosphorylation and inhibition of polycystic kidney disease 1 protein (PKD1) (Sumara et al. 2009). PKD1 regulates insulin secretion and cell survival, and mice lacking p38δ are protected against stress-induced β cell death and have elevated insulin secretion.
The p38 MAPK family regulates TNFα levels, a key cytokine in the control of systemic inflammation. While p38α regulates TNFα expression by controlling its transcription and mRNA stability (Campbell et al. 2004, Kang et al. 2008), recent studies demonstrate key roles for p38γ and p38δ in the posttranscriptional regulation of TNFα biosynthesis, with myeloid-specific p38γ/δ gene ablation triggering defects in TNFα production (Risco et al. 2012, Gonzalez-Teran et al. 2013). Translational elongation of nascent pro-TNFα protein is mediated by eukaryotic elongation factor 2 (eEF2) kinase (Gonzalez-Teran et al. 2013), which is inhibited by p38γ/δ-mediated phosphorylation (Knebel et al. 2001). Neutrophils are an important cell type that triggers inflammation and monocyte migration (Nathan 2006) and that makes an important contribution to liver steatosis development and metabolic dysfunction (Talukdar et al. 2012, Bertola et al. 2013, Mansuy-Aubert et al. 2013). p38γ/δ have also been shown to control their recruitment and mobility (Ittner et al. 2012, Gonzalez-Teran et al. 2013). However, the specific contribution of these kinases to chronic inflammation induced by metabolic disease has not been tested.
Several reports have demonstrated that SAPKs control several metabolic events and their excessive activation is associated with deleterious effects during obesity. Most of these studies have examined JNK family members, in particular the isoforms JNK1 and JNK2. Further research is needed to clarify the function of the different p38 isoforms during obesity and in metabolic regulation and to define the cell types in which these kinases perform their function. The functions of p38γ/δ have not been studied extensively in vivo, and very little is known about their substrates and functions. Recent reports suggest that the four isoforms of p38 have widely differing functions (Sumara et al. 2009, Gehart et al. 2010, Sabio et al. 2010b, Ittner et al. 2012, Gonzalez-Teran et al. 2013, Sabio & Davis 2014). Studies in knockout and conditionally deficient mouse models for each p38 kinase are needed in order to define the role of these proteins in health and disease.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the review.
G S is investigator of the Ramón y Cajal Program and recipient of a Principe de Girona award. E M is a fellow of the Caixa program. This work was funded by the following grants to G S: ERC 260464, EFSD 2030, MICINN SAF2013-43506-R and Comunidad de Madrid S2010/BMD-2326. The Centro Nacional de Investigaciones Cardiovasculares (CNIC) is supported by the Spanish Ministry of Economy and Competitiveness and the Pro-CNIC Foundation.
We thank S Bartlett for English editing.
CampbellJCiesielskiCJHuntAEHorwoodNJBeechJTHayesLADenysAFeldmannMBrennanFMFoxwellBM2004A novel mechanism for TNFα regulation by p38 MAPK: involvement of NF-κB with implications for therapy in rheumatoid arthritis. Journal of Immunology1736928–6937. (doi:10.4049/jimmunol.173.11.6928).
CaoWDanielKWRobidouxJPuigserverPMedvedevAVBaiXFloeringLMSpiegelmanBMCollinsS2004p38 mitogen-activated protein kinase is the central regulator of cyclic AMP-dependent transcription of the brown fat uncoupling protein 1 gene. Molecular and Cellular Biology243057–3067. (doi:10.1128/MCB.24.7.3057-3067.2004).
ImbernonMBeiroaDVazquezMJMorganDAVeyrat-DurebexCPorteiroBDiaz-ArteagaASenraABusquetsSVelasquezDA2013Central melanin-concentrating hormone influences liver and adipose metabolism via specific hypothalamic nuclei and efferent autonomic/JNK1 pathways. Gastroenterology144636–649. (doi:10.1053/j.gastro.2012.10.051).
IttnerABlockHReichelCAVarjosaloMGehartHSumaraGGstaigerMKrombachFZarbockARicciR2012Regulation of PTEN activity by p38δ-PKD1 signaling in neutrophils confers inflammatory responses in the lung. Journal of Experimental Medicine2092229–2246. (doi:10.1084/jem.20120677).
Juge-AubryCEHammarESiegrist-KaiserCPerninATakeshitaAChinWWBurgerAGMeierCA1999Regulation of the transcriptional activity of the peroxisome proliferator-activated receptor α by phosphorylation of a ligand-independent trans-activating domain. Journal of Biological Chemistry27410505–10510. (doi:10.1074/jbc.274.15.10505).
KanetoHNakataniYKawamoriDMiyatsukaTMatsuokaTAMatsuhisaMYamasakiY2005Role of oxidative stress, endoplasmic reticulum stress, and c-Jun N-terminal kinase in pancreatic β-cell dysfunction and insulin resistance. International Journal of Biochemistry & Cell Biology371595–1608. (doi:10.1016/j.biocel.2005.04.003).
KimJKGimenoREHigashimoriTKimHJChoiHPunreddySMozellRLTanGStricker-KrongradAHirschDJ2004bInactivation of fatty acid transport protein 1 prevents fat-induced insulin resistance in skeletal muscle. Journal of Clinical Investigation113756–763. (doi:10.1172/JCI200418917).
Mansuy-AubertVZhouQLXieXGongZHuangJYKhanARAubertGCandelariaKThomasSShinDJ2013Imbalance between neutrophil elastase and its inhibitor α1-antitrypsin in obesity alters insulin sensitivity, inflammation, and energy expenditure. Cell Metabolism17534–548. (doi:10.1016/j.cmet.2013.03.005).
PradaPOZecchinHGGasparettiALTorsoniMAUenoMHirataAECorezola do AmaralMEHoerNFBoscheroACSaadMJ2005Western diet modulates insulin signaling, c-Jun N-terminal kinase activity, and insulin receptor substrate-1ser307 phosphorylation in a tissue-specific fashion. Endocrinology1461576–1587. (doi:10.1210/en.2004-0767).
RiscoAdel FresnoCMambolAAlsina-BeauchampDMacKenzieKFYangHTBarberDFMorcelleCArthurJSLeySC2012p38γ and p38δ kinases regulate the toll-like receptor 4 (TLR4)-induced cytokine production by controlling ERK1/2 protein kinase pathway activation. PNAS10911200–11205. (doi:10.1073/pnas.1207290109).
SolinasGVilcuCNeelsJGBandyopadhyayGKLuoJLNauglerWGrivennikovSWynshaw-BorisAScadengMOlefskyJM2007JNK1 in hematopoietically derived cells contributes to diet-induced inflammation and insulin resistance without affecting obesity. Cell Metabolism6386–397. (doi:10.1016/j.cmet.2007.09.011).
TsaousidouEPaegerLBelgardtBFPalMWunderlichCMBronnekeHCollienneUHampelBWunderlichFTSchmidt-SupprianM2014Distinct roles for JNK and IKK activation in agouti-related peptide neurons in the development of obesity and insulin resistance. Cell Reports91495–1506. (doi:10.1016/j.celrep.2014.10.045).
ZhangXXuAChungSKCresserJHSweeneyGWongRLLinALamKS2011Selective inactivation of c-Jun NH2-terminal kinase in adipose tissue protects against diet-induced obesity and improves insulin sensitivity in both liver and skeletal muscle in mice. Diabetes60486–495. (doi:10.2337/db10-0650).