Metabolic inflammation is a very topical area of research, wherein aberrations in metabolic and inflammatory pathways probably contribute to atherosclerosis, insulin resistance (IR) and type 2 diabetes. Metabolic insults arising from obesity promote inflammation, which in turn impedes insulin signalling and reverse cholesterol transport (RCT). Key cells in the process are metabolically activated macrophages, which up-regulate both pro- and anti-inflammatory pathways in response to lipid spillover from adipocytes. Peroxisome proliferator-activated receptors and AMP-activated protein kinase (AMPK) are regulators of cellular homeostasis that influence both inflammatory and metabolic pathways. Dietary fats, such as saturated fatty acids (SFAs), can differentially modulate metabolic inflammation. Palmitic acid, in particular, is a well-characterized nutrient that promotes metabolic inflammation via the NLRP3 (the nod-like receptor containing a pyrin domain) inflammasome, which is partly attributable to AMPK inhibition. Conversely, some unsaturated fatty acids are less potent agonists of metabolic inflammation. For example, monounsaturated fatty acid does not reduce AMPK as potently as SFA and n-3 polyunsaturated fatty acids actively resolve inflammation via resolvins and protectins. Nevertheless, the full extent to which nutritional state modulates metabolic inflammation requires greater clarification.
In recent times, the relationship between metabolism and immune function has come to the fore. There is now a greater understanding of nutritional and metabolic regulation of immune function and its impact on whole-body homeostasis. Metabolically triggered inflammation or ‘metabolic inflammation’ is a complex multicellular and interorgan process which probably contributes to the pathogeneis and progression of obesity, IR, type 2 diabetes (T2D) and cardiovascular disease (CVD) . Metabolic inflammation impedes insulin signalling and reverse cholesterol transport (RCT), the process by which cholesterol is excreted from the body . Metabolic inflammation is accentuated by obesity, which, by virtue of its escalating prevalence, highlights the importance of this complex biological process . Nutrients, such as palmitic acid, oleic acid, eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), can positively or negatively influence metabolic inflammation through various mechanisms. Conversely, weight loss and exercise may attenuate metabolic inflammation. Here, we discuss recent developments in these areas, specifically the impact of fatty acid composition on macrophage function, the resolution of inflammation, insulin signalling and RCT.
Metabolic inflammation, triggered by certain nutrients or metabolic stress, is characterized by the activation of pro-inflammatory signalling pathways and cytokine production in metabolic tissues, e.g. in adipose tissue, during obesity  (Figure 1). In the normal state, excess fat is typically stored in adipocytes as triacylglycerol (TAG), cholesterol and cholesteryl esters in a safe non-lipotoxic environment. In the obese state, excessive TAG accumulation and cholesterol imbalances can result in metabolically unhealthy hypertrophic adipocytes which are insulin-resistant with impaired differentiation potential [4,5]. These stressed adipocytes produce pro-inflammatory adipokines, including tumour necrosis factor alpha (TNF-α), monocyte chemoattractant protein-1 and interleukin (IL)-6, as well as leaking lipids which become metabolic stressors . Adipose tissue macrophages (ATMs) infiltrate adipose tissue to clear these excess lipids. ATMs produce pro-inflammatory cytokines, such as IL-1β, TNF-α and IL-6, which further propagate inflammation  (Figure 1).
The ‘metabolic’ macrophage
ATMs act as accessory cells, clearing excess lipids and phagocytizing apoptotic and necrotic stressed adipocytes . ATMs constitute up to 40% of all obese adipose tissue cells, compared with ∼10% of all lean adipose tissue cells . Classically, activated macrophages utilize ‘aerobic glycolysis’, or the metabolism of glucose to lactate in the presence of abundant oxygen, as their main fuel source . Interestingly, preliminary data from our group suggest that macrophages treated with palmitic acid utilize glycolysis to a larger extent, whereas oleic acid maintains cells in the more energy efficient oxidative phosphorylative state. Metabolically activated ATMs, or those activated with SFA such as palmitic acid, possess a phenotype that is mechanistically distinct from classical immune macrophage activation . Kratz et al. demonstrate that treating ATMs with palmitic acid activates both inflammatory and metabolic pathways. On the one hand, palmitic acid binds to Toll-like receptor 4 (TLR4), which activates the transcription factor nuclear factor (NF)-κB, resulting in the secretion of pro-inflammatory cytokines such as TNF-α and IL-1β (Figure 2). On the other hand, palmitic acid internalization by ATMs leads to activation of the lipid-responsive transcription factor peroxisome proliferator-activated receptor gamma (PPARγ) (Figure 2). PPARγ is a lipid-responsive transcription factor that is highly expressed in adipose tissue where it plays a major role in adipocyte differentiation, lipid homeostasis and triglyceride storage . Contrastingly, PPARγ activation suppresses NF-κB activity, thus regulating the balance between cytokine production and lipid metabolism .
The NLRP3 inflammasome
The NLRP3 (the nod-like receptor containing a pyrin domain) inflammasome is a major instigator of metabolic inflammation that is activated by cell stress and fatty acids including palmitate. It tightly regulates IL-1β and IL-18 cytokine production by a two-step mechanism. The first priming state involves activation of the innate immune receptor TLR4 by lipopolysaccharide (LPS) or SFA, which results in the transcription of an inactive pro-form of IL-1β or IL-18 [11,12]. A second signal related to mitochondrial damage and cell stress results in the formation of a protein complex which activates the protease caspase-1. Caspase-1 then cleaves pro-IL-1β or pro-IL-18 into active secreted forms . Vandanmagsar et al. demonstrated that weight loss in obese individuals with T2D is associated with reduced expression of NLRP3 in AT, as well as with decreased inflammation and improved insulin sensitivity. Ablation of NLRP3 in obese mice also results in improved insulin sensitivity . Palmitic acid induces IL-1β-mediated inflammation through TLR4 signalling, leading to activation of the NLRP3 inflammasome . Our group has shown that replacing SFA with monounsaturated fatty acids (MUFAs) or polyunsaturated fatty acids (PUFAs) is an effective strategy for dampening of NLRP3-induced inflammation via modulation of the energy sensor AMP-activated protein kinase (AMPK). AMPK is a key metabolic regulator that senses excess energy in the form of intracellular ATP and is activated by the anti-inflammatory salicylates [15,16]. Replacement of SFA with MUFA within a high-fat diet (HFD) maintained AMPK signalling and presented reduced NLRP3-mediated inflammation despite equivalent obesity, resulting in improved insulin sensitivity . Similarly, our group has also demonstrated the ability to modulate inflammation using a compound derived from milk-casein, which acts by inhibiting both steps of NLRP3 inflammasome activation . The contribution of NLRP3 to inflammation in metabolic tissues has recently been extensively reviewed .
The resolution of inflammation
The resolution of inflammation is an active metabolic inflammatory process involving a reduction in neutrophil number, increased monocyte recruitment, stimulation of macrophage uptake of apoptotic neutrophils and clearance of phagocyte by the lymphatic system . Many specialized pro-resolving mediators such as ‘anti-inflammatory’ nutrients may enable this resolution. The long-chain n-3 PUFAs (LC n-3 PUFAs), EPA and DHA, are precursors to lipid mediators resolvins and protectins, each of which actively alleviate and/or resolve inflammation . For example, Lipoxin A4, Resolvin E1 and Protectin D1 all stimulate non-inflammatory phagocytosis of apoptotic neutrophils by macrophages and can reduce secretion of TNF-α, IL-1β and IL-12 from various other immune cells . Previous work by our group demonstrated the ability of DHA, and to a lesser extent EPA, to reduce LPS-induced NF-κB activation and TNF-α secretion in macrophages . Our group also demonstrated the pro-resolving effects of lipoxins through inhibition of TNF-α and interferon (IFN)-γ expression, while stimulating pro-resolving IL-10, in a model of early renal fibrosis . In a genetically induced obese mouse model, DHA supplementation also demonstrated beneficial effects including increased adiponectin (an anti-inflammatory and insulin-sensitizing adipokine), induced AMPK phosphorylation and up-regulated PPARα, resulting in improved glucose transport and insulin receptor signalling .
However, translation of these beneficial effects into human interventions has presented inconsistent results. This is potentially due to differences in the dose of fatty acids administered, baseline health status of the participants, and genetic and potentially other unknown determinants of response . Examples of some papers that illustrate efficiency in man include Itariu et al. , who demonstrated modulation of systemic inflammation and adipose tissue in severely obese patients using LC n-3 PUFA. The LC n-3 PUFA group presented decreased IL-6 and TAG levels, decreased expression of inflammatory genes in subcutaneous adipose tissue and increased production of anti-inflammatory eicosanoids . Furthermore, combined supplementation of LC n-3 PUFA with plant sterols resulted in decreased C-reactive protein (CRP), TNF-α, IL-6 and increased adiponectin levels in hyperlipidemic participants, with a 22.6% reduction in CVD risk . Skulas-Ray et al. compared a nutritional dose of EPA and DHA with a pharmaceutical dose in healthy participants with moderate hypertriglyceridaemia. The pharmaceutical dose lowered TAG levels by 27%. There was no difference in cholesterol, endothelial function or markers of inflammation with either dose over an 8-week period . However, a 4-month intervention in overweight males presented reductions in IL-6 and TNF-α levels with two doses of LC n-3 PUFA; therefore, duration of supplementation is a potential limiting factor in the anti-inflammatory potential . The recommended intake of EPA and DHA (250 mg/day) is much lower than the doses administered. A diet high in α-linolenic acid decreased CVD risk by reducing serum CRP, vascular cell adhesion protein (VCAM)-1, e-selectin, total cholesterol, low-density lipoprotein and TAG . Consumption of conjugated linoleic acid-enriched dairy products did not, however, affect inflammatory markers, insulin sensitivity or cholesterol . In summary, current results would suggest that the anti-inflammatory impact of LC n-3 PUFA is established in vitro, and in many animal models of inflammation. However, the translation of the effect into humans is dependent on both the dose and the health status of the individual, highlighting the importance of a personalized nutrition approach.
Similiarly, Kien et al.  demonstrated in young healthy adults that lowering the SFA/MUFA ratio of a diet over 3 weeks resulted in reduced secretion of IL-1β, IL-18, IL-10 and TNF-α in stimulated peripheral blood mononuclear cells and also decreased circulating TNF-α levels. Previous research by our group replaced dietary SFA with either MUFA or LC n-3 PUFA for 12 weeks in obese participants with metabolic syndrome, and failed to observe differences in insulin sensitivity, cholesterol, inflammatory markers or blood pressure. This may be due to the unhealthy baseline status of the participants or the irreversible effects of a high-SFA diet . The differences in the type of SFA consumed is important, as not all SFAs have detrimental effects. SFAs with an even number of carbon atoms in their chain (myristic, palmitic and stearic acid) are associated with a higher risk of T2D . Alcohol intake, or the replacement with dietary fat with carbohydrate, leads to the mobilization of these even-numbered fats from fat stores. SFAs with an odd number (pentadecylic and margaric acid), commonly found in dietary and fermented dairy products, are associated with a lower risk of T2D .
Insulin signalling regulates glucose, lipid and energy homeostasis, mainly acting in metabolic tissues, i.e. liver, skeletal muscle and adipose tissue. During obesity, ‘lipid spillover’ from expanding adipose tissue drives inflammation in metabolic tissues . Excess free fatty acids accumulate in these tissues, resulting in lipotoxicity and an increase in potentially harmful intracellular lipid products such as diacylglycerols and ceramides . SFAs impair insulin signalling through TLR4, activating the NF-κB pathway and NLRP3 inflammasome . In this way, increased free fatty acid levels inhibit insulin signalling, resulting in insulin resistance (IR). The inhibition of hepatic glucose production by insulin and stimulation of glycogen synthesis are impaired, and accumulation of free fatty acids in the liver results in hepatic steatosis . High glucose combined with high levels of SFAs, such as palmitic acid and stearic acid, can result in pancreatic β-cell toxicity, further exacerbating the hyperglycaemic state . Thus, by having an impact on metabolic-inflammation, SFAs also influence insulin signalling. The contribution of metabolic inflammation to IR in metabolic tissues has recently been extensively reviewed .
Reverse cholesterol transport
RCT is the process by which cholesterol is transported to the liver for excretion . In peripheral tissues, macrophages efflux cholesterol via ATP-binding cassette (ABC) transporters ABCA1 and ABCG1 onto pre-β- and mature high-density lipoprotein (HDL) particles, respectively . These HDL particles deliver the cholesterol load to the liver where they may be excreted via bile or faeces . In the metabolically stressed state, inefficient excretion of HDL particles can lead to a build-up of cholesterol, increasing the risk of atherosclerosis or fatty liver disease . Inflammation inhibits cholesterol efflux from macrophages to plasma, leading to the cellular accumulation of cholesterol . In an obese mouse model, RCT was increased with a high-MUFA diet; however, this was impeded by high-SFA diets. Both SFA and MUFA diets had increased macrophage-to-plasma efflux, and plasma-to-liver efflux; however, cholesterol was trapped in the liver with increased body weight with the SFA group. This was attributed to pro-inflammatory markers found on smaller HDL particles and reduced expression of cholesterol transporters ABCG5 and ABCB11 in the liver . The effect of LC n-3 PUFA from fish oil on RCT has also been investigated. Soybean oil (mixture of fats), hydrogenated coconut oil (higher SFA) and menhaden oil (higher PUFA) were given to mice for 4 weeks before RCT was measured using radiolabelled cholesterol-loaded macrophages. The high-SFA diet demonstrated reduced macrophage-to-plasma efflux compared with a high-PUFA or -MUFA diet. The high-PUFA diet had increased liver-to-faeces RCT and reduced cholesterol in the liver compared with the other HFDs . Similar results were obtained in a study in golden Syrian hamsters fed a low-PUFA or high-PUFA diet. Both studies demonstrated increased macrophage-to-faeces RCT and increased hepatic mRNA expression of ABCA1/ABCG5 [44,45]. Thus, the replacement of SFA for MUFA or supplementation with PUFA may increase the latter steps of RCT through up-regulation of hepatic cholesterol transporters. However, the mechanisms underlying how different metabolic stressors regulate ABCA1- and ABCG1-mediated cholesterol efflux and expression are still lacking.
Tweaking metabolic inflammation via weight loss and exercise
A novel approach in developing our understanding of the resolution of metabolic inflammation is to decipher mechanisms of lifestyle interventions such as caloric restriction, weight loss and exercise on metabolic inflammation in obese individuals at risk of T2D. This is highly pertinent given the close biological relationship between metabolism, inflammation and insulin sensitivity. Ruffino et al.  demonstrated that moderate exercise leads to reduced inflammatory markers in macrophages, which in turn lead to improved insulin sensitivity, possibly though PPARγ activation. Another potential mechanism is by production of the ketone β-hydroxybutyrate, which blocks activation of the NLRP3 inflammasome through inhibition of caspase-1 activation, independently of AMPK or glycolytic mechanisms . Progressive weight loss is another model suitable for deciphering mechanisms involved in the resolution of metabolic inflammation. Magkos et al. carried out a study comparing the effects of moderate (5%) and progressive (11–16%) weight loss on metabolic and inflammatory markers. Both weight losses resulted in decreased intra-abdominal adipose tissue volume, intrahepatic TAG content and systolic blood pressure. A reduction in TAG preceded progressive improvements in adipose tissue biology, inflammatory markers, up-regulation of metabolic pathways and genes involved in cholesterol flux and down-regulation of metabolic pathways and genes involved in lipid synthesis, ECM remodelling and oxidative stress. This gives insights into the stepwise improvements in the resolution of inflammation towards a normal phenotype .
Clearly, dysregulation of highly interlinked metabolic and inflammatory pathways during obesity leads to obesity-associated complications including IR, T2D and CVD. Modulation of these pathways through diet, exercise and anti-inflammatory interventions is an attractive therapeutic approach for these complications. However, the exact mechanism involved in the resolution of this meta-inflammation needs to be elucidated. Our work illustrating the paradigm wherein different fatty acids modulate the interplay between metabolic and inflammatory pathways provides proof of concept. However, translating this to man and the development of nutritional interventions to combat dysregulated metabolic inflammatory circuits in obesity require further classification.Figure 1,Figure 2
Consequences of obesity and associated metabolic inflammation.
Interplay between adipocytes and macrophages in the obese state.
AMP-activated protein kinase
adipose tissue macrophages
free fatty acid
- LC n-3 PUFA
long-chain n-3 polyunsaturated fatty acids
monounsaturated fatty acids
the nod-like receptor containing a pyrin domain
peroxisome proliferator-activated receptors
polyunsaturated fatty acids
reverse cholesterol transport
saturated fatty acids
type 2 diabetes
tumour necrosis factor
vascular cell adhesion molecule
H.M.R. is the recipient of Science Foundation Ireland (SFI) principal investigator award [11/PI/1119]. A.M.K. and H.M.R. are supported by Enterprise Ireland [TC2013-0001]. Y.M.L., M.E.O. and H.M.R. are funded by the Irish Department of Agriculture, Food and the Marine, ‘Healthy Beef’ Programme [13/F/514]. F.C.M. is jointly funded by SFI, the Health Research Board (HRB) and the Wellcome Trust [097311/Z/11/Z] under the SFI–HRB–Wellcome Trust Biomedical Research Partnership.
The Authors declare that there are no competing interests associated with the manuscript.