Mechanisms of obesity and related pathology: linking immune responses to metabolic stress


K. P. Karalis, Foundation for Biomedical Research of the Academy of Athens, 4 Soranou Efessiou Street, GR-11527 Papagou, Greece
Fax: +30 210 6597545
Tel: +30 210 6597465


There is a tightly regulated interaction, which is well-conserved in evolution, between the metabolic and immune systems that is deranged in states of over- or under-nutrition. Obesity, an energy-rich condition, is characterized by the activation of an inflammatory process in metabolically active sites such as adipose tissue, liver and immune cells. The consequence of this response is a sharp increase in circulating levels of proinflammatory cytokines, adipokines and other inflammatory markers. Activation of the immune response in obesity is mediated by specific signaling pathways, with Jun N-terminal kinase and IκB kinase β/nuclear factor κ-light-chain-enhancer of activated B cells being the most well studied. It is known that the above events modify insulin signaling and result in the development of insulin resistance. The nutrient overload characterizing obesity is a metabolic stressor associated with intracellular organelle (e.g. the endoplasmic reticulum) stress. The exact characterization of the series of events and the mechanisms that integrate the inflammatory response with metabolic homeostasis at the cellular and systemic level is a very active research field. In this minireview, we discuss the signaling pathways and molecules associated with the development of obesity-induced inflammation, as well as the evidence that supports a critical role for the stress response in this process.


11β-hydroxysteroid dehydrogenase


corticotropin-releasing hormone


endoplasmic reticulum


IκB kinase β




Jun N-terminal kinase


nuclear factor κ-light-chain-enhancer of activated B cells


neuropeptide Y


peroxisome proliferator-activated receptor


Toll-like receptor


tumor necrosis factor-α


Obesity is the epidemic of our century, with sharply and steadily rising rates [1,2]. The ongoing rise in obesity rates reflects changes in lifestyle, including nutritional and other environmental challenges, that are superimposed on susceptible genetic background. Various genetic studies, ranging from an identification of the cause(s) of monogenic obesity [3–6] to the latest genome-wide association studies, support a genetic contribution in the development of this disease [7–9].

The major adverse consequences of obesity, insulin resistance and the corresponding type 2 diabetes and atheromatosis, when added together, account for a great number of disease-related deaths. If the obesity-related cancer cases are added to this number, obesity-related mortality by far exceeds that of other common diseases. The latter indicates the urgent need to develop novel efficient therapeutic modalities for this condition. The common denominator in the pathogenesis of the co-morbidities of obesity is the presence of an active inflammatory process in tissues that are important for metabolism, such as adipose, liver, muscle and endothelium. Epidemiological evidence has related obesity and type 2 diabetes with inflammation for more than 100 years [10–13]. Clinical data have shown impaired numbers of natural killer (NK), T cells and neutrophils in obese patients [14,15], although the molecular mechanisms underlying this association only started to be revealed in the late 1990s [16,17].

Obesity impairs systemic metabolic homeostasis and, as such, it elicits a stress response [18], whereas stress has been linked to the development of obesity and more specifically visceral obesity [19]. Stress is characterized by increasing levels of glucocorticoid, a steroid hormone necessary for the development and differentiation of preadipocytes [20]. Further support of the role of stress and glucocorticoid in the development of obesity came from studies on 11β-hydroxysteroid dehydrogenase (11β-HSD), an enzyme that catalyzes the conversion of inactive cortisol (cortisone) to active 11β-HSD1 or vice versa (i.e. active cortisol to inactive 11β-HSD2). The expression of the enzyme is tissue-specific and controls, in a tightly regulated manner, the local actions of glucocorticoid [21]. Protection from diet-induced obesity and insulin resistance was found in mice lacking 11β-HSD1 [22,23], whereas mice with adipose tissue-specific overexpression of 11β-HSD1 developed an obese phenotype when fed a high-fat diet [24]. Studies in obese humans resulted in similar findings [25,26] and, most importantly, suggested that the ability to regulate the activity of 11β-HSD is lost in type 2 diabetic patients, whereas it is compromised in nondiabetic, obese individuals [26]. These findings indicate a dynamic regulation of the activity of 11β-HSD that is gradually lost as the severity of the disease increases and progresses to the development of the metabolic syndrome and type 2 diabetes. The above support findings the rationale for the therapeutic application of inhibitors of this enzyme in progressing obesity in humans [27].

The physiological responses mobilized under stressful situations form the adaptive stress response that aims to re-establish equilibrium. The hypothalamic–pituitary–adrenal axis and the central and peripheral components of the autonomic nervous system constitute the two main vital stress-system functions [28]. Chronic stress represents a prolonged threat to homeostasis that can progressively lead to complications caused by both the stressor and the prolongation of the adaptive response. The overnutrition-induced chronic stress disturbs the balance between metabolic and immune functions and contributes to the development of visceral obesity, type 2 diabetes, atherosclerosis and the metabolic syndrome. The associated secretion of proinflammatory cytokines by the adipose tissue may act as an additional chronic stimulus for activation of the hypothalamic–pituitary–adrenal axis. All the above promote the maintenance of a vicious cycle between metabolic and immune responses in states of nutrient overload, such as in obesity, resulting in a chronic active inflammatory condition.

The association of inflammation with obesity has been indicated by epidemiological studies from the 1950s onwards, although the underlying mechanisms of this process remained unknown for several decades. A major contribution in our understanding of the inflammatory nature of obesity was the identification of tumor necrosis factor-α (TNFα) expressed in the adipose tissue of experimental animals [29,30]. A series of studies followed these initial findings of little more than one decade ago and revealed impaired insulin action after the administration of TNFαin vitro, in cells, and in vivo, in mice [31]. Furthermore, obese mice with genetically impaired TNFα expression or function had significantly improvement in insulin resistance [32].

The discovery of the local production of a major proinflammatory cytokine in the fat and, furthermore, its ability to regulate insulin action systemically gave a new impetus to this field. Soon it was found that adipose tissue is the site of production of several proinflammatory factors such as interleukin (IL)-6 and monocyte chemotactic protein-1 [33,34]. Circulating levels of these factors are elevated in obese patients as well as other inflammation-related molecules including C-reactive protein, serum amlyloid A, fibrinogen and plasminogen activator inhibitor-1 [35,36]. It soon became apparent that, in obese states, adipose tissue is infiltrated by macrophages that are partially responsible for the inflammatory changes described above [37,38]. The compromised ability of macrophages for chemotaxis and infiltration of the adipose tissue resulted in a significant decrease of insulin resistance and the corresponding development of atheromatosis [39]. It was also found that, in obesity, there is a remarkable shift in the pool of tissue macrophages from the alternatively-activated M2 type to the classically-activated M1 type [40,41]. This phenomenon results in changes in secreted cytokines from predominantly anti-inflammatory (M2) to proinflammatory (M1) [41,42]. Alternatively-activated macrophages protect from the development of obesity-related inflammation and insulin resistance, most likely by regulating peroxisome proliferator-activated receptor (PPAR)γ activity [41]. The exact mechanisms underlying this effect and the contribution of the shift in the cytokines profile according to the predominance of alternatively versus classically-activated macrophages has not yet been fully clarified. Adipocytes and macrophages share common functions, such as phagocytosis, and great similarities in their transcriptional profile. It has been proposed that the preadipocyte differentiation process is characterized by plasticity, which is expressed by their ability to transdifferentiate [43–45]. Finally, adipocytes from different body depots may have major differences in their inflammatory phenotype, with visceral fat expressing the more ‘malignant’ phenotype and subcutaneous fat the most ‘benign’ phenotype. The beneficial effects of the ablation of specific fat depots in the reversal of insulin resistance and inflammation have confirmed the above theory, although the steps preceding the acquirement of these specific characteristics are not yet known [46].

The secretion of proinflammatory factors by adipose tissue and, most importantly, the regulation of their secretion by increasing adiposity substantiated the hypothesis for an ongoing low-grade inflammatory process during obesity. Furthermore, the identification of similar changes in the expression of these factors in the systemic circulation of patients with obesity provided further incentives for continuing research on the mechanisms that underlie the obesity-related development of inflammation. Innate immune receptors such as Toll-like receptor (TLR) 4 and TLR2 are expressed in the adipose tissue and their expression is induced in obese individuals [47,48]. It was found that free fatty acids binding innate immune receptors such as TLR4 are directly linked to the development of inflammation in states of hyperlipidemia, such as obesity. Binding of free fatty acids to TLR4 led to induction of the cascade of events resulting in the activation of the ΙκB/nuclear factor κ-light-chain-enhancer of activated B cells (NFκB) system and Jun N-terminal kinase (JNK) and to the subsequent induction of proinflammatory factors [49]. In support of the significance of this process, Tlr4 deficient mice did not exhibit any signs of inflammation subsequent to diet-induced obesity [50–52]. Recently it was shown that the anti-inflammatory effects of statins in the adipose tissue are also mediated by locally expressed innate immune receptors [53].

In addition to the proinflammatory cytokines that are expressed in fat, but may originate from cells other than adipocytes, such as the infiltrating macrophages or stromal cells, specific cytokines collectively named adipokines are produced from adipocytes. In 1994, Zhang et al. [54] isolated a hormone, leptin, from the fat, a discovery that changed our thinking on the biological role of this tissue in the maintenance of energy homeostasis. Thus, it became apparent that adipose tissue is not a lipid storage site alone, but is involved in the regulation of energy homeostasis and communicates actively with the other participants in this process, such as the liver, muscle and hypothalamus. The main function of leptin is to signal the energy state of the organism acting through specific receptors expressed in several tissues apart from the adipose. Obesity is characterized by leptin resistance that promotes the vicious cycle of ‘regulation’ of energy homeostasis associated with this condition. Leptin has also immunomodulatory properties mediated primarily by lymphocytes and are exerted at a systemic level [55,56]. The direct contribution of leptin resistance in the development of obesity-associated inflammation has not yet been elucidated. Subsequent to the discovery of leptin, other adipokines, with resistin, visfastin and adiponectin being the most abundantly expressed, were identified [57–59]. Circulating levels of adipokines in humans reflect the degree of their adiposity. Immunomodulatory, pro- or anti-inflammatory, properties have been described for adipokines, although it is not clear how these changes in expression are related to the development of the inflammation of obesity.

Adipose tissue expresses neuropeptides and/or their specific receptors. Thus, angiotensin [60], neuropeptide Y (NPY) and NPY receptors are expressed in the adipose tissue [61,62]. It was suggested that NPY acting on the adipose tissue is involved in the development of the metabolic syndrome [63]. Functional orexin receptors have been demonstrated in adipose tissue because stimulation with orexins resulted in the activation of PPARγ and the inhibition of hormone sensitive lipase [64]. Furthermore, neuropeptides with demonstrated immunomodulatory properties have been identified in the adipose tissue, with their expression increasing in parallel to the expansion of this tissue, such as in obesity. Substance P and neurokinin receptor 1 are expressed in the adipose tissue as well as in the mesenteric fat in cases of inflammation in the adjacent intestinal tissue [65]. Corticotropin-releasing hormone (CRH), a major mediator of the stress response secreted by the paraventricular nucleus of the hypothalamus [66], is also secreted at inflammatory sites and possesses potent proinflammatory properties that influence innate and acquired immune processes [67]. We have recently shown CRH expression in the adipose tissue in levels that are significantly increased in diet-induced obese mice compared to lean mice (K. Karalis, E. Kokkotou, Y. Koutmani & L. van Vlerken, unpublished observation). Expression of CRH receptor 1 in human adipocytes and the regulation of gene expression in human adipocytes in culture by the specific CRH receptor 1 antagonist, antalarmin, was reported previously [68,69], whereas CRH induced lipogenesis in cultured sebocytes [70]. The biological role of immunomodulatory neuropeptides has been demonstrated in other inflammatory diseases, such as intestinal inflammation [64,65]. Elucidation of their role in the adipose tissue may result in the development of novel targeted therapies that aim to control the development of obesity-induced inflammation and insulin resistance.

Mechanisms linking obesity to inflammation

Obesity is a chronic low-grade inflammatory disease characterized by the eased expression of proinflammatory cytokines and gradually worsening insulin sensitivity. Inhibition of insulin signaling downstream of the insulin receptor occurs primarily by phosphorylation of insulin-receptor substrate-1 at inhibitory serine residues instead of the stimulatory tyrosine residues [71]. It remains unclear whether insulin resistance precedes the development of inflammation or vice versa. In obese mice, the activity of JNK, a member of the mitogen-activated protein kinase family otherwise referred as mitogen-activated protein kinase 8, is increased and, upon association with insulin-receptor substrate-1, results in phosphorylation of Ser307 and impaired insulin signaling in the adipose and other metabolically active tissues, such as the liver [72]. The significance of the latter was appreciated when mice genetically deficient in JNK1, one of the three JNK isoforms, were protected from the diet-induced development of insulin resistance and diabetes [73]. Furthermore, experimental induction of JNK was sufficient to induce diabetes. Activated JNK induces the expression of proinflammatory cytokines such as TNFα and IL-6 [74]. JNK is activated by changes in homeostasis caused by various stimuli, including proinflammatory cytokines, increasing circulating levels of lipids, reactive oxygen species and infectious stimuli [75].

Other factors shown to link obesity to inflammation include IκB kinase β (IKKβ), PNC, GSK3β, S6K, Erk and mTOR, whereas nuclear receptors such as PPARs and liver X receptors have been shown to inhibit the activation of inflammatory pathways [76]. Specifically, activation of IKKβ results in induction of NFκB, a transcription factor with a critical role in the upregulation of proinflammatory factors. Blockade of the activation of NFκB in obese mice by treatment with sodium salicylate reversed insulin resistance [17], whereas similar trials in human patients have provided promising results so far. Furthermore, the inhibition of NFκB [77] or JNK [78] in macrophages resulted in the reversal of diet-induced insulin resistance.

Inflammatory signaling pathways are also triggered by lipids. Free fatty acids bind innate immune receptors such as TLR4 and, thus, are directly linked to the development of inflammation in states of hyperlipidemia, such as obesity. Binding of free fatty acids to TLR4 led to the subsequent induction of the cascade of events, resulting in the activation of the ΙΚΚβ/NFκB system and the induction of proinflammatory cytokines [49]. In addition, Tlr4 deficient mice did not exhibit any signs of inflammation subsequent to diet-induced obesity. Expression of TLR2 and TLR4 was increased in the adipose tissue of obese individuals, whereas it was shown that locally expressed innate immune receptors mediated the anti-inflammatory effects of statins in adipose tissue [53]. Transport of the fatty acids inside the cells requires their interaction with fatty acid binding proteins, a family of proteins expressed in several tissues, including adipose tissue and macrophages. Members of this family have been associated with the development of diabetes, insulin resistance and inflammation. Ongoing studies aim to evaluate the exact role of fatty acid binding proteins in immune and other cells, whereas antagonists of these factors have been tested for therapeutic actions in obesity, diabetes and atheromatosis [76,79,80].

In obesity, the cells and the intracellular organelles are exposed to increased stress conditions mainly as a result of metabolic overload [76]. More specifically, mitochondria and the endoplasmic reticulum (ER) appear to be the most sensitive organelles to metabolic stressors. The development of hypoxic conditions in the expanded adipose tissue during obesity results in an increased production of reactive oxygen species and the corresponding development of oxidative stress that activates JNK and other kinases, as described above [81]. ER becomes overloaded by lipids and its ability to address the increasing demands through the unfolded protein response gradually declines [82]. Activation of ER stress results in the activation of JNK and the corresponding induction of proinflammatory cytokines via different pathways involving inositol requiring kinase 1, pancreatic ER kinase and activating transcription factor 6 [83,84]. The contribution of the activated ER stress in the development of insulin resistance was demonstrated by the reversal of insulin resistance after the administration of chaperones that blocked this activation [85,86]. Recently, chaperones administered intracerebroventricularly in the hypothalamus of ob/ob mice reversed the obesity-associated leptin resistance [87], supporting the putative therapeutic application of these compounds in obesity and insulin resistance.

The proinflammatory mediators, intracellular processes and signaling pathways linking obesity to inflammation have been identified and their effects were confirmed in experimental animal models. In humans, several similarities with the animal findings have been found, as described above. Interestingly, attempts for therapeutic interventions based on the mouse studies, such as neutralization of TNFα, have not been successful so far. Obese humans administered TNFα monoclonal antibodies did not show similar responses to those of mice with respect to a decrease in body weight and the reversal of insulin resistance. Ongoing studies aim to evaluate the efficacy of IL-1 antagonists [88] and NFκB inhibitors, such as sodium salsalate [89], in the treatment of obesity and insulin resistance. The usual dilemma in multigenic diseases is the relevance of the experimental data (derived in the majority of cases from mouse studies) to human disease. The above described human studies targeting TNFα in obese patients have not supported the initial enthusiasm elicited by the mouse data. On the other hand, studies in humans with monogenic obesity have confirmed the findings obtained in mice, supporting the similarity in the underlying pathology of the disease in the two species. More studies are needed to characterize and highlight the exact similarities and differences between the mechanisms operating in humans and mice, in order to take full advantage of the vast amount of available experimental data in combating the human disease. Along these lines, two recent independent studies using human and rodent adipose and liver tissues were designed to identify the gene networks that are affected by susceptibility loci and are also associated with the development of obesity [90,91]. Both studies unmasked a network enriched in factors associated with inflammatory and immune responses causally associated with obesity-related traits. These findings confirm the plethora of experimental studies suggesting that obesity is an inflammatory disease and support the rationale for the therapeutic application of antiinflammatory compounds in obese patients, as discussed above. Finally, one very important message of these studies is that a systems biology approach in complex diseases, focusing on specific gene networks rather than single factors, makes the possibile development of successful personalized pharmacological treatments a realistic target.