Why do anti-inflammatory therapies fail to improve insulin sensitivity?

Abstract

Chronic inflammation occurs in obese conditions in both humans and animals. It also contributes to the pathogenesis of type 2 diabetes (T2D) through insulin resistance, a status in which the body loses its ability to respond to insulin. Inflammation impairs insulin signaling through the functional inhibition of IRS-1 and PPARγ. Insulin sensitizers (such as rosiglitazone and pioglitazone) inhibit inflammation while improving insulin sensitivity. Therefore, anti-inflammatory agents have been suggested as a treatment strategy for insulin resistance. This strategy has been tested in laboratory studies and clinical trials for more than 10 years; however, no significant progress has been made in any of the model systems. This status has led us to re-evaluate the biological significance of chronic inflammation in obesity. Recent studies have consistently asserted that obesity-associated inflammation helps to maintain insulin sensitivity. Inflammation stimulates local adipose tissue remodeling and promotes systemic energy expenditure. We propose that these beneficial activities of inflammation provide an underlying mechanism for the failure of anti-inflammatory therapy in the treatment of insulin resistance. Current literature will be reviewed in this article to present evidence that supports this viewpoint.

Introduction

For about two decades, it has been known that inflammation contributes to obesity-associated insulin resistance. Inflammatory cytokines (eg, TNF-alpha, IL-1, and IL-6) have been shown to induce insulin resistance in multiple organs (fat, muscle and liver). TNF-α elevation was found in adipose tissue of obese mice in 1993¹. That study provided the first evidence of the role of chronic inflammation during obesity and its association with insulin resistance in an animal model. Macrophages in adipose tissue are the major source of inflammatory cytokines in obesity^{2, 3}. Recent studies from multiple groups, including ours, consistently suggest that adipose tissue hypoxia is a root of chronic inflammation in obesity⁴. Hypoxia is likely the result of a reduction in blood flow to adipose tissue, which is supported by some studies in humans and animals^{5, 6, 7}.

In addition to adipose tissue hypoxia, metabolites of fatty acids and glucose, including diacylglyceride (DAG), ceramide, and reactive oxygen species, also contribute to the chronic inflammation in obesity. They activate the inflammatory response in several ways. They can directly interact with signaling kinases (PKCs, JNKs, and IKKs) in cells⁸; the lipids can also signal through cell membrane receptors for lipids, such as TLR4, CD36, or GPR^{8, 9, 10, 11, 12, 13}. Fat or glucose oxygenation in the mitochondria can also generate reactive oxygen species (ROS), which can then induce activation of the inflammatory kinases (JNK and IKK) in the cytoplasm. The lipids also induce endoplasmic reticulum (ER) stress to activate JNK and IKK^{14, 15}. In obesity, these signaling pathways are activated as a result of the surplus calories and involved in the pathogenesis of chronic inflammation.

Chronic inflammation and insulin resistance

At the molecular level, inflammation induces insulin resistance by targeting IRS-1 and PPARγ.

Inflammation and IRS-1 (insulin receptor substrate 1)

In cellular models of insulin resistance, the pro-inflammatory cytokine, TNF-α, is widely used to induce insulin resistance. The data from these cellular studies suggest that TNF-α is a major risk factor for insulin resistance in obesity and other chronic diseases^{1, 16, 17}. TNF-α inhibits insulin signaling by serine phosphorylation of IRS-1, which leads to the dissociation of IRS-1 from the insulin receptor and causes degradation of IRS-1 protein^{17, 18, 19}. In the insulin signaling pathway, IRS-1 undergoes tyrosine in response to insulin stimulation, which leads to activation of the insulin signaling pathway, downstream PI3K/Akt activation, and Glut4 translocation to the cell membrane for glucose uptake. TNF-α induces insulin resistance by IRS-1 serine phosphorylation through the activation of several serine kinases, including JNK^{20, 21}, IKK²², ERK^{23, 24, 25}, PKC^{26, 27, 28}, Akt^{28, 29}, GSK-3^{30, 31, 32}, IRAK³³, and mTOR^{34, 35}. In a recent study, we showed that IKK2 (IKKβ) inhibits IRS-1 function through the activation of S6K, which directly phosphorylates IRS-1 at multiple sites (such as S312/307 and S270/265) in TNF-α-treated cells^{22, 36}. Serine phosphorylation induces IRS-1 degradation and serves as a negative feedback signal to impair insulin action³⁵.

Inflammation inhibits PPARγ function

The IKKβ/NF-κB (nuclear factor kappa B) pathway is a dominant inflammatory signaling pathway. The pathway has been under active investigation in the obesity field after IKKβ was found to induce insulin resistance in obese mice³⁷. The serine kinase IKK has three major isoforms, including IKKα (IKK1), IKKβ (IKK2), and IKKγ, which requires IKKβ for NF-κB activation³⁸. In obesity, IKKβ is activated by several intracellular signals, such as ROS, ER stress, DAG, and ceramide. IKKβ is also activated by extracellular stimuli, including TNF-α, IL-1, fatty acids¹¹ and hypoxia³⁹. IKKβ induces NF-κB activation by phosphorylation of the Inhibitor of Kappa B alpha (IκBα)⁴⁰.

NF-κB is a ubiquitous transcription factor that is formed by two subunits of the Rel family, which includes seven members, p65 (RelA), p50 (NF-κB1), c-Rel, RelB, p100, p105, p52⁴¹. These members form a homodimer or heterodimer that regulates gene transcription. In most cases, NF-κB is a heterodimer of p65 and p50. P65 contains the transactivation domain and mediates the transcriptional activity of NF-κB. P50 inhibits the transcriptional activity of p65⁴², and the NF-κB activity is enhanced in p50 knockout mice⁴³. NF-κB inhibits PPARγ function through the competition for transcriptional coactivators or the exchange of corepressors with PPARγ⁴⁴. This process is responsible for inhibiting PPAR-target genes, such as CAP and IRS-2. Our study shows that IKK promotes the activity of HDAC3 in the nuclear corepressor complex. IKK induces nuclear translocation of HDAC3 from the cytoplasm. In the cytosol, HDAC3 associates with IκBα, and the degradation of IκBα promotes HDAC3 translocation into the nucleus. The PPARγ inactivation leads to suppression of IRS-2 expression, a signaling molecule in insulin signaling pathways for Glut4 translocation.

Free fatty acids and insulin resistance

Elevated plasma free fatty acids (FFAs) induce insulin resistance in obese and diabetic subjects⁴⁵. It was known as early as 1983 that lipid infusion caused insulin resistance^{46, 47}. To examine the mechanism by which FFAs induced insulin resistance in vivo, rats were tested in a hyperinsulinemic-euglycemic clamp after a 5-h infusion of lipids/heparin, which raises plasma FFA concentrations⁴⁷. FFAs resulted in an approximate 35% reduction in insulin sensitivity, indicated by the glucose infusion rate (P<0.05 vs control), and a 25% reduction in glucose transport activity, as assessed by 2-[1,2-3H]deoxyglucose uptake in vivo (P<0.05 vs control). PKCθ is a major kinase involved in FFA-induced insulin resistance⁴⁸. According to the Randle glucose-fatty acid cycle, the preferential oxidation of free fatty acids over glucose plays a major role in the pathogenesis of insulin sensitivity⁴⁹. Local accumulation of fat metabolites, such as ceramides, diacylglycerol or acyl-CoA, inside skeletal muscle and liver may activate a serine kinase cascade, leading to defects in insulin signaling and glucose transport⁵⁰.

Inflammation and energy metabolism

Inflammation is associated with increased energy expenditure in patients with chronic kidney disease⁵¹, cachexia⁵², inflammatory bowel disease⁵³ and Crohn's disease⁵⁴. NF-κB activity can promote energy expenditure, as supported by documents on energy expenditure in cachexia^{55, 56} and infection. However, the role of NF-κB in energy expenditure was not tested in transgenic models. To this end, we have investigated energy metabolism in transgenic mice with elevated NF-κB activity. The transcriptional activity of NF-κB is enhanced either by over-expression of NF-κB p65 in the fat tissue, or inactivation of NF-κB p50 by global gene knockout^{57, 58}. In these two models, inflammatory cytokines (TNF-α and IL-6) were elevated in the blood, and energy expenditure was increased both during the day and at night^{57, 58}. Expression of TNF-α and IL-1 mRNA was increased in adipose tissue and macrophages. These cytokines are positively associated with energy expenditure in the body⁵⁶. In transgenic mice with deficiencies in these cytokines or their receptors, energy accumulation is enhanced and energy expenditure is reduced. This positive energy balance has been reported in transgenic mice deficient in TNF-α⁵⁹, IL-1⁶⁰, or IL-6⁶¹.

The above literature suggests that energy accumulation induces chronic inflammation. Inflammation may promote energy expenditure in a feedback manner to counteract an energy surplus⁶². Inflammation may act in the peripheral organs/tissues, as well as in the central nervous system, to regulate energy balance. In the peripheral tissues, inflammation induces fat mobilization and oxidation to promote energy expenditure. In the central nervous system, inflammation can inhibit food intake and activate neurons for energy expenditure, while inhibition of inflammation leads to fat accumulation⁶².

Anti-inflammation therapies for insulin resistance

In clinical trials, high-dose salicylate was used to inhibit inflammation by targeting IKK/NF-κB^{37, 63, 64, 65}. Salicylate reduces blood glucose by inhibiting IKK/NF-κB, as seen decades ago in patients with diabetes^{64, 65, 66}. More studies demonstrated that high-doses of aspirin (∼7.0 g/d) improved multiple metabolic measures in patients with T2D, including substantial reductions in fasting and postprandial glucose, triglycerides and FFAs. These changes were associated with reduced hepatic glucose production and improvements in insulin-stimulated glucose disposal, assessed during hyperinsulinemic-euglycemic clamping^{63, 64, 65, 67}. Aspirin inhibits the activity of multiple kinases induced by TNF-α, such as JNK, IKK, Akt, and mTOR. It may enhance insulin sensitivity by protecting the IRS proteins from serine phosphorylation⁶⁸. However, the therapeutic value of high-dose aspirin is limited by its side effects, including gastrointestinal irritation and high risk of bleeding.

Statins, a class of anti-inflammatory drugs, have been shown to downregulate the transcriptional activity of NF-κB, AP-1, and HIF-1α^{65, 69}, with coordinated reductions in the expression of prothrombotic and inflammatory cytokines. Randomized clinical trials have demonstrated that statins reduces CRP, multiple cytokines, and inflammatory markers in the body. Even with modest anti-inflammatory properties, statins do not appear to enhance insulin resistance or significantly improve glycemia⁷⁰. A recent review published in JAMA suggests that statin therapy is associated with excess risk for diabetes mellitus. The researchers analyzed five earlier trials, involving 32 752 patients, to test the effect of the drug dose. Those getting intensive treatment were 12 percent more likely to have diabetes⁷¹, which translates into a 20 percent increase in developing diabetes in the high-dose statin users compared to those who do not take the drugs.

Glucocorticoids are the most effective anti-inflammatory drugs used to treat inflammatory diseases. Dexamethasone is a potent synthetic member of the glucocorticoid class of steroid drugs. In a clinical study, the effect of dexamethasone on insulin-stimulated glucose disposal was investigated with a double-blind, placebo-controlled, cross-over trial comparing insulin sensitivity (measured by the euglycemic hyperinsulinemic clamp) in young healthy males allocated the placebo or 1 mg dexamethasone twice daily for 6 d, each in random order. Six days of dexamethasone therapy was associated with a 30% decrease in insulin sensitivity^{72, 73}. This indicates that strong inhibition of inflammation may block the beneficial effects of inflammation on insulin sensitivity.

Interleukin-1β induces inflammation in islets of patients with type 2 diabetes⁷⁴. The interleukin-1–receptor antagonist, a naturally occurring competitive inhibitor of interleukin-1⁷⁵, protects human beta cells from glucose-induced functional impairment and apoptosis⁷⁶. The expression of the interleukin-1-receptor antagonist is reduced in pancreatic islets of patients with type 2 diabetes mellitus. High glucose induces the production of interleukin-1β in human pancreatic beta cells, leading to impaired insulin secretion, decreased cell proliferation, and enhanced apoptosis. In this double-blind, parallel-group trial involving 70 patients with type 2 diabetes⁷⁴, 34 patients were randomly assigned to receive 100 mg of anakinra (a recombinant human interleukin-1-receptor antagonist) subcutaneously once daily for 13 weeks. In the control group, 36 patients received placebo. All patients underwent an oral glucose-tolerance test. At the end of the study, the two study groups exhibited no difference in insulin resistance, insulin-regulated gene expression in skeletal muscle, serum adipokine levels, and the body-mass index. However, the therapy did improve blood glucose levels. The authors conclude that the improvement is from enhanced pancreatic β-cell function. This study indicates that inhibition of IL-1β improves glucose metabolism, independent of insulin sensitivity.

TNF-α expression is elevated in the adipose tissue of obese rodents and humans. In animal studies, administration of exogenous TNF-α induced insulin resistance, whereas neutralization of TNF-α improved insulin sensitivity. TNF-α knockout mice were used to examine the role of TNF-α in obesity-associated insulin resistance⁷⁷. The KO mice were compared with WT mice in lean and obese (induced by gold-thioglucose [GTG]-injection) conditions at 13, 19, and 28 weeks of age. In the lean condition, the KO mice exhibited a 14% reduction in body weight at 28 weeks of age. The epididymal fat pad was decreased by 25% in weight, relative to those of the wild-type littermate controls. Fasting glucose was reduced slightly by 10%, but the glucose response in an oral glucose tolerance test (OGTT) was not affected. In the obese condition, the body weight was identical between the KO and WT mice. Glucose levels were significantly increased in both groups during the OGTT. This indicates that the absence of TNF-α is not sufficient to protect mice from insulin resistance in obese conditions⁷⁷. Some animal studies⁷⁸ and several clinical trials using TNF antagonism have thus far failed to improve insulin sensitivity^{79, 80, 81, 82, 83}. These facts suggest that there are many unknowns in the relationship of obesity-associated inflammation and insulin resistance.

The role of IL-6 in the pathogenesis of obesity and insulin resistance is controversial. IL-6 knockout (KO) mice were compared with WT littermate mice in lean or obese conditions. IL-6 KO mice displayed obesity, hepatosteatosis, liver inflammation and insulin resistance when compared with the lean condition on a standard chow diet⁸⁴. Overexpression of IL-6 was also used to test insulin resistance in mice. In the study, IL-6 overexpression was generated in skeletal muscle, and the IL-6 protein levels were increased in the circulation. The mice lost both body weight and body fat in response to IL-6 in this model, even though their food intake remained unchanged⁸⁵. These observations suggest that IL-6 increases energy expenditure. In the IL-6 mice, insulin levels were elevated, and hypoglycemia was observed⁸⁵. In another study, Sadagurski et al demonstrated that a high level of IL-6 in the circulation reduces obesity and improves metabolic homeostasis in vivo⁸⁶.

The role of the anti-inflammatory cytokine IL-10 has been studied in the pathogenesis of obesity and insulin resistance⁸⁷. IL-10 is a critical cytokine of M2 (type 2) macrophages. A recent study has identified the roles of M1 (pro-inflammatory) and M2 (anti-inflammatory) macrophages in the regulation of insulin sensitivity⁸⁸. An increase in M2 macrophages and a decrease in M1 macrophages within the adipose tissue are associated with enhanced insulin sensitivity. In another study, the hematopoietic-cell-restricted deletion of IL-10 in mice was used to study the relationship between IL-10 and insulin resistance⁸⁹. The mice were assessed for insulin sensitivity in an insulin tolerance test in lean (chow diet) and obese (high fat diet) conditions. The results show that deletion of IL-10 from the hematopoietic system does not have an effect on insulin resistance⁸⁹. Other studies suggest that IL-10 cannot improve insulin sensitivity in diet-induced obese mice or humans^{90, 91}.

New potential drug candidates for insulin resistance

The antidiabetic drug thiazolidinedione (TZD) restores insulin action by activating PPARγ, thus lowering the levels of FFAs in the blood. Activation of PPARγ improves insulin sensitivity in rodents and humans through a combination of metabolic actions, including partitioning of lipid stores and regulating metabolic and inflammatory mediators, termed adipokines⁹². However, TZD-based medicines for insulin sensitization have many side effects: troglitazone (Rezulin) was associated with massive hepatic necrosis; rosiglitazone (Avandia) and muraglitazone, with increased cardiovascular events; and now, pioglitazone has been associated with bladder cancer⁹³. These adverse events suggest that the thiazolidinedione-based drugs may not be safe in the long-run. It is necessary to discover a new class of drug to treat insulin resistance.

Recent studies indicate that histone deacetylase (HDAC) inhibitors may be a new class of drug candidates for insulin sensitization. HDACs are key enzymes in regulating gene expression. Protein acetylation is one type of epigenetic regulation of gene expression. Acetylation is controlled by histone acetyltransferases (HATs) and histone deacetylases (HDACs). Histone acetylation by HATs opens the chromatin structure to activate gene transcription, while histone deacetylases (HDACs) repress gene expression. HDACs are divided into three classes: class I HDACs (1, 2, 3, 8, 11), class II HDACs (4, 5, 6, 7, 9, 10)⁹⁴ and class III HDACs (SIRT1-7)⁹⁵. Inhibition of histone deacetylase activity has been reported as a new approach to treat diabetes mellitus^{96, 97, 98}. In our study, supplementation of histone deacetylase inhibitors, butyrate or Trichostatin A, prevented high-fat diet-induced obesity and improved insulin sensitivity in mice. HDAC inhibition promoted energy expenditure, and reduced blood glucose and triglyceride levels in mice⁹⁸. HDAC inhibits insulin resistance on a molecular level by the following means: a) reducing the lipid toxicity^{44, 99, 100, 101, 102}; b) reducing chronic systemic inflammation^{103, 104, 105, 106, 107, 108}; c) promoting beta-cell development, proliferation, differentiation and function⁹⁷; and d) promoting energy expenditure^{98, 109}. Based on their multiple beneficial effects, HDAC inhibitors may represent a novel drug in the treatment of insulin resistance. However, clinical trials are needed to test this concept.

Conclusions

Type 2 diabetes is one of the major diseases associated with obesity. It is known that obesity promotes type 2 diabetes through insulin resistance, a state in which bodies lose their responsiveness to insulin. Many studies confirm that inflammation and free fatty acids (FFAs) are major pathogenic factors for insulin resistance in obese conditions. The most effective therapy for insulin resistance is to reduce both FFA and inflammation. Diminishing inflammation by anti-inflammatory drugs does not significantly improve insulin sensitivity in animal models or in clinical trials because inflammation is beneficial in regulating energy metabolism. Inhibiting this beneficial activity is likely to cause the failure of anti-inflammatory drugs in treating insulin resistance. Current literature consistently reports that fatty acids remain a therapeutic target in the treatment of insulin resistance. As an insulin sensitization-drug, TZD reduces both FFA and inflammation in the body. However, TZDs have many side effects such as obesity, heart attacks, and bladder cancer. HDAC inhibitors may be a new class of drug for treating insulin resistance by promoting energy expenditure and preventing obesity.