Expansion of adipose tissue is accompanied by chronic low-grade inflammation that primes target organs for the development of obesity-associated chronic inflammatory diseases. Adipose tissue-resident immune cells play a major role in the induction and regulation of obesity-induced systemic inflammation. These can be proinflammatory immune cells (e.g., neutrophils, dendritic cells, M1 macrophages, Th1 cells, B cells, and mast cells) as well as anti-inflammatory immune cells (e.g., regulatory T cells, Th2 cells, M2 macrophages, and eosinophils). Although most types of immune cells are already present in the adipose tissue, their number increases dramatically with the progression of obesity.
Granulocytes in Adipose Tissue Inflammation
Neutrophils present fundamental mechanisms of effector cells (e.g., opsonization, agglutination, complement activation, regulation of inflammation) and participate in initiation of immune response and resolution of inflammation. Low circulating adiponectin level characteristic to obesity was shown to induce neutrophil activity and number in the peripheral blood [
41]. Activated neutrophils infiltrate adipose tissue early during diet-induced obesity in mice in an attempt to limit the local inflammatory process [
42]. Moreover,
in vitro studies have shown that neutrophils physically bind adipocytes in a CD11b/ICAM-1 interaction and in a manner dependent on their activation state [
42]. A recent study evidenced that diet-induced obesity in mice determined a rapid increase in adipose tissue’s neutrophil presence, lasting up to 90 days, and a parallel increased expression in the activity of neutrophil elastase [
42]. Neutrophil elastase seemed to influence the following macrophage infiltration and M1 polarization, since M2 (alternatively polarized) macrophages were prevalent in obese mice lacking this enzyme [
43].
Mast cells are important sensors of acute inflammation triggered by pathogenic bacteria and also play an important part in allergic type reactions [
44]. More recent evidence implicates these cells in cardiometabolic diseases [
45]. Ironically, when Paul Ehrlich described them in 1878, he coined them “Mastzellen” or “fattening” cells based on their granule-enriched cytoplasm. Mast cells share a common bone marrow precursor with basophil granulocyte. Both cell types respond to IgE stimulation following an allergen encounter, and they release similar mediators responsible for local and systemic anaphylactic reaction [
46,
47]. As opposed to basophils, mast cells leave the bone marrow in an immature state and then fully differentiate in specific tissue sites. Thus, mast cells display tissue specificity and are more intimately related to specific homeostatic and pathologic states. Mast cells respond to microenvironment by releasing preformed contents of granules (histamine, heparin, tryptase, and chymase) or
de novo synthesis of proinflammatory cytokines such as IL-6, IL-8, and TNF-α. Based on protease content, we distinguish either tryptase or tryptase/chymase-expressing mast cells. In terms of localization, mast cells are found in two main compartments: mucosal surfaces and perivascular connective tissue. Mast cells grow and proliferate in response to growth factors, stem cell factor (SCF), and nerve cell growth factor (NGF) as well as cytokines (IL-3, IL-4, IL-9, IL-10). Abnormal expansion of visceral adipose tissue is accompanied by influx of immune cells. Mouse models of diet-induced obesity showed accumulation of mast cells in adipose tissue [
45,
48,
49]. Mast cell Kit
W-sh/W-sh-deficient mice, lacking only mature mast cells, are resistant to diet-induced obesity and are able to maintain glucose homeostasis when fed with a high-fat diet. Analysis of their visceral adipose tissue revealed a significant reduction in proinflammatory cytokines and chemokines [
48] and a decrease in macrophage number. Therefore, it appears that mast cell arrival in adipose tissue precedes the release of proinflammatory mediators that attract macrophages. Furthermore, even the pharmacological inhibition of mast cell degranulation reproduced the metabolic phenotype seen in Kit
W-sh/W-sh-deficient mice.
Human adipose tissue appears to contain both tryptase and tryptase/chymase-expressing mast cells. Despite similar representation in both lean and obese subjects, it appears that mast cells in the latter group have an increased rate of degranulation [
50]. Moreover, obese subjects that progressed to complications like diabetes were found to have a higher number of mast cells. Visceral fat mast cells from obese patients were found to produce significantly higher proinflammatory cytokines (IL-1, IL-6) and macrophage chemoattractant (MCP-1) previously shown to induce insulin resistance [
51,
52]. Adipose tissue fibrosis has been linked to obesity insulin resistance and abnormal cytokine/adipokine secretion from adipose cells [
45,
53]. Development of obesity in diabetic
db/
db mice was associated with recruitment of immature mast cells and whose maturation paralleled the metabolic abnormalities. Mast cell-derived tryptase was found to promote collagen 5 mRNA expressions in fibroblasts and was associated to adipose tissue fibrosis in
db/
db mice [
45]. Antifibrotic compounds (tranilast, angiotensin-converting enzyme inhibitors, and silymarin) coupled with dietary interventions could prevent mast cell maturation and degranulation to reduce associated metabolic abnormalities [
53].
Recently, the effects of adipose tissue eosinophils have also been documented on local macrophage activity and polarization. In the adipose tissue, alternative (M2) activation of macrophages is driven by the cytokine interleukin-4 (IL-4). Eosinophils are the major IL-4-expressing cells in white adipose tissues of mice. In their absence, the M2 macrophage number is greatly attenuated leading to impaired glucose tolerance and insulin resistance [
54]. Thus, recent studies suggest that beyond monocytes and macrophages, plenty of other myeloid cells, such as dendritic cells, lymphoid cells like NK cells, NKT cells, B and T lymphocytes, and eosinophils, could play a combined role in the inflammatory process associated with obesity. Due to the presence of such an immune cell spectrum, several researchers consider adipose tissue as an ancestral lymphoid organ where physiologic and pathologic immune processes can take place simultaneously [
55].
Dendritic Cells, Monocytes, and Macrophages in Adipose Tissue Inflammation
Dendritic cells (DCs) are specialized, heterogeneous group of mononuclear cells able to acquire, process, and present antigens to naïve T cells. Based on their phenotype and functional characteristics, DCs can be found in almost all tissues and are further divided into the following: conventional/myeloid DCs (CD11
+), plasmacytoid DCs (CD11c
−CD303
+), and a novel group of inflammatory DCs (inf-DCs) generated from
in situ activation of monocytes recruited into the site of inflammation. Several studies consider obesity-induced adipose tissue hypoxia and elevated level of plasma free fatty acids (FFAs) as potential initiating events in the activation and recruitment of DCs into the enlarged adipose tissue. Bertola et al. [
56] showed for the first time the accumulation of specific inflammatory dendritic cells CD11c
highF4/80
low in the adipose tissue of obese mice and CD11c
+CD1c
+ in the adipose tissue of obese patients. The emergence and expansion of CD11c
highF4/80
low DCs the in obese mice and CD11c
+CD1c
+ in the obese patients induced proinflammatory Th17 cell responses and macrophage accumulation and correlated with higher BMI and insulin resistance. Mice lacking DCs had a reduced number of macrophages in the adipose tissue, whereas DC replacement in DC
−/−mice increased macrophage populations in the adipose tissue. Moreover, lean wild-type mice that received bone marrow-derived DCs had macrophage infiltration in the adipose tissue, while mice lacking DCs completely were resistant to the high-fat diet weight gain and metabolic abnormalities [
57]. Importantly, Hagita et al. [
58] proved that adipose tissue location can dictate the degree of associated vasculature inflammation. In an
in vivo study, they showed that mice that had lean visceral fat transplanted around the femoral artery presented increased vascular inflammation (leukocyte and DC recruitment to the femoral artery) as compared to mice that had lean subcutaneous fat transplanted around femoral artery. Moreover, when they are used for transplantation, the visceral/subcutaneous fat from donor mice fed with a high-fat diet, the inflammatory response at the femoral artery level was substantially increased. Therefore the effect of high-fat diet on adipocytes is compartment specific [
58].
New studies have shown that in response to high-fat diet, the hypertrophied adipocytes produce more CCL20, a chemoattractant whose receptor—CCR6—is highly expressed on adipose tissue dendritic cells. In addition, the adipose tissue dendritic cells express higher levels of IL-6, TGF-β, and IL-23 [
59]. These are essential cytokines for Th17 cell proliferation and differentiation. Co-cultures of adipose tissue dendritic cells and naïve T cells promoted proinflammatory Th17 cell differentiation and IL-17 production. This effect was significantly increased when compared with dendritic cells derived from spleen [
59]. These studies show that adipose tissue DCs are among the first to sustain the expanded adipose inflammatory milieu. Furthermore, by recruiting and activating other immune cells, including monocytes and macrophages, the adipose tissue DCs propagate the immune response associated with adipose tissue expansion [
57].
Monocytes are also heterogeneous for phenotype and function, and different subsets rise in response to microenvironment cues. Two main monocyte subsets may be distinguished based on their expression of specific receptors: in humans CD14 (LPS receptor) and CD16 (FcgammaRIII) and in mice Ly6C and Gr1. Based on the relative expression of CCR2 and CX3C chemokine receptor 1 (CX3CR1), Ly6C
hi monocytes are Gr1
+CCR2
+CX3CR1
lo and correspond to human CD14
++CD16
− (classical monocytes) whereas Ly6C
lo monocytes are Gr1
−CCR2
−CX3CR1
hi and correspond to human CD14
+CD16
+ (nonclassical monocytes). Circulating nonclassical monocytes demonstrate a patrolling behavior along blood vessel walls [
60] and form “standby” deposits in noninflamed peripheral tissues such as spleen, lung, and liver [
61].
However, despite the overall conservation, the comparison of the two species’ subsets highlighted some diversity such as expression of fatty acid translocase (FAT/CD36), tetraspanin CD9, triggering receptor expressed on myeloid cells 1 (Trem-1), and PPARγ. Recently, Shantsila et al. [
62] demonstrated unequivocally that human monocyte group includes three major functionally and phenotypically different subsets: the classical CD14
+CD16
−CCR2
+, the intermediate CD14
+CD16
+CCR2
+, and the nonclassical CD14
dimCD16
+CCR2
− monocytes.
Most of the monocytes are CD14
+CD16
− and can amount to up to 85 % in healthy subject [
62]. The CD16
+ monocytes increase their frequency in response to chronic inflammatory conditions, such as chronic kidney disease (CKD) [
63], obesity [
64,
65], and associated cardiovascular diseases [
66]. High levels of the CD14
+CD16
+ subset of CD16
+ were associated with cardiovascular events [
67] and reduced survival at 35 months in CKD patients [
68]. In the same time, CD14
dimCD16
+ subtype was positively correlated with the BMI [
65] and atherogenic lipoproteins and inversely associated with high-density lipoprotein cholesterol.
Poitou and colleagues [
69] investigated the frequency of CD16
+ monocyte subsets and their potential role in obesity and weight loss; they showed an increase in CD14
dimCd16
+ monocytes in obese and diabetic patients. Importantly, weight loss as well as surgery-induced weight loss caused a reduction of CD14
dimCD16
+ monocytes that correlated with reduction of subclinical atherosclerosis, as evaluated by intima–media thickness.
Obesity promotes the mobilization of monocytes from the bone marrow in part by activating the CCR2. Deficiency of CCR2 or its ligand, MCP-1, in mice results in failure of monocyte mobilization and is associated with protection from monocyte infiltration into adipose tissue and insulin resistance [
70,
71]; Spite et al. [
72] report that activation of the leukotriene B4 (LTB4) and its receptor BLT-1 axis is required for obesity-induced increases in peripheral blood monocytes and subsequent adipose tissue macrophage accumulation.
Adipose tissue macrophages are the main component of adipose tissue immune cells (40–60 % of all adipose tissue immune cells), and their number increases progressively after only 1 week of high-fat diet feeding [
73]. Two major macrophage phenotypes have been described: classically activated or M1, which trigger a proinflammatory, type 1 immune response, and alternatively activated or M2, which promote anti-inflammatory, type 2 responses during the healing process [
74,
75]. However,
in vivo, monocytes and macrophage phenotypes more likely represent points on a spectrum with high plasticity that shapes obesity-induced inflammation.
Progressive adipose tissue expansion is accompanied by macrophage accumulation and decreased expression of key genes of adipocyte differentiation (PPARγ and C/EBPα). This reduces the number of new, small adipocyte recruitment and leads to mature adipocyte hypertrophy [
76]. Nevertheless, there are different rates of macrophage accumulation based on the anatomical location of the adipose tissue (i.e., visceral vs subcutaneous fat). Human subcutaneous adipose tissue macrophages retrieved by liposuction from healthy, overweight women are composed mainly of cells expressing CD206, a marker of activated macrophages. However, it seems that only the CD206
+/CD16
+ cells accumulate in the adipose tissue directly proportional with adiposity. Although a rapid local differentiation of inflammatory monocytes into macrophages cannot be excluded, enhanced local proliferation might be involved in the accumulation of CD206
+/CD16
+ cells.
Significant differences in MCP-1 production and in the amount of infiltrated macrophages were found in the subcutaneous, epididymal, renal, and mesenteric fat samples from obese and control mice. The MCP-1 protein levels were significantly higher in the obese mice than those in the nonobese controls, with the highest MCP-1 level detected in the mesenteric adipose tissue sample from obese mice. Moreover, the differences in MCP-1 level among anatomically different adipose tissues correlated with the number of macrophages infiltrated into that fat pad. These results indicate that the mesenteric adipose tissue is a major depot for MCP-1, which can modulate macrophage trafficking and activation during obesity-related inflammatory diseases [
77]. On the other hand, an experimental study in mice evidenced new players such as microRNAs (miR-233) that suppress the classic proinflammatory (M1) pathways and enhance the alternative anti-inflammatory (M2) responses in the adipose tissue [
78].
Natural Killer Cells and Natural Killer T Cells in Adipose Tissue Inflammation
Obesity is accompanied by a low-grade, systemic inflammatory process that involves both innate and adaptive immunity.
High-fat diet feeding stimulates the secretion of interferon gamma (IFN-γ) in the adipose tissue of wild-type mice. Several studies showed that IFN-γ initiates early accumulation of T and B lymphocytes in the adipose tissue and activates local macrophage recruitment and their classical M1 differentiation [
79]. In humans, O’Rourke et al. [
80] showed that visceral adipose tissue from obese individuals presented elevated IFN-γ transcript levels and a high frequency of macrophages, T cells, and natural killer (NK) cells relative to subcutaneous adipose tissue. On the other hand, obese but IFN-γ-deficient mice had significantly less adipose tissue expression of inflammatory genes such as TNF-α and MCP-1 and better glucose tolerance than the obese, control mice consuming the same diet [
81]. Moreover, the absence of T and B lymphocytes in the RAG2
−/−mice fed with a high-fat diet had no effect on the increased macrophage accumulation in the expanded adipose tissue or insulin resistance [
79]. In conclusion, in the absence of T and B cells, the NK cells are also able to produce IFN-γ and TNF-α, which are relevant to macrophage recruitment in the adipose tissue during obesity.
The early role of natural killer T (NKT) cells and their regulation in adipose tissue immune response are not yet thoroughly deciphered. Currently, it is known that NKT cells can bridge innate and adaptive immune responses [
82]. There are three types of NKT cells: invariant NKT (iNKT), noninvariant NKT, and NKT-like cells. Invariant NKT and noninvariant NKT cells are CD1d dependent [
83]. The MHC class I-like CD1d glycoprotein is a member of the CD1 family of antigen-presenting molecules and is responsible for the selection of NKT cells. Importantly, NKT cell population and CD1d expression was found to be highly expressed in adipocytes from obese mice and humans compared to those from lean mice and lean human subjects [
84]. In addition, the CD1d-expressing adipocytes are able to stimulate NKT cell activity through mere physical interaction. In animal studies, CD1d
(−/−) mice fed with a high-fat diet gained little weight, had less liver inflammation, and presented smaller adipocytes in comparison with wild-type control mice on the same diet [
85]. The NKT cell-deficient Jα18
−/−mouse model fed with a high-fat diet became more obese and displayed increased adipose tissue inflammation in the early stage of obesity. These results underline the role of NKT in the early adipose tissue inflammation and obesity-related insulin resistance [
73,
84].
T cells in Adipose Tissue Inflammation
Adaptive immunity seems to assign a causative role to B and T lymphocytes in activating innate immunity [
86], while regulatory T (Treg) cells have a suppressive function, rescuing obese mice from chronic adipose tissue inflammation [
87].
Duffaut et al. [
79,
88] observed that fat expresses a predominant macrophage population with CD3
+-activated T cells (including CD4
+ T helper and CD8
+ T cytotoxic cells), a minor number of CD56 NK cells, and few CD19
+ B lymphocytes. The CD3
+CD56
+ NKT cells and CD25
+ Treg cells were found in a very low number in steady state. Interestingly, most CD3
+-activated T cells were organized in clusters surrounding adipocytes, and their number increased proportionally with the adipose tissue size and BMI. However, this distribution was influenced by the degree of obesity and by adipose tissue location. Visceral adipose tissue from obese patients showed an increased number of macrophages and lymphocytes, especially CD8
+ effector T cells, compared to subcutaneous fat. Moreover, proinflammatory chemokines followed a similar pattern and increased proportionally to the amount of visceral adipose tissue [
79,
88]. Both CD4
+ and CD8
+ T cells have been found in adipose tissue, and their number increases with obesity [
81] in both humans and mice [
89]. Depletion of CD8
+ cells in obese mice decreased the number of macrophages in adipose tissue and lowered TNF-α and IL-6 levels, while T cell receptor (TCR)
−/−mice were clearly protected against obesity-induced hyperglycemia and insulin resistance [
90]. Accordingly, adoptive transfer of CD8
+ cells induced M1 macrophage accumulation, impaired glucose tolerance, and insulin sensitivity in obese mice [
86].
Fabbrini et al. showed that adipose tissue from insulin-resistant obese patients had 3- to 10-fold more CD4
+ T cells that produced IL-22 and IL-17 in comparison with adipose tissue from insulin-sensitive obese and lean subjects. IL-17 and IL-22 inhibited uptake of glucose through receptors for IL-22 and IL-17 expressed in the human liver and skeletal muscle [
91].
Treg cells are thought to maintain tolerance/anti-inflammatory microenvironment through IL-10 production. Tregs are abundant in visceral adipose tissue of lean mice, but their number is significantly reduced in insulin-resistant mice models of obesity [
87]; similarly, a reduced number of Foxp3
+ Treg cells was found in visceral adipose tissue from obese humans [
89]. The signal transducer and activator of transcription 3 (STAT3) plays an important role in the Th1/Treg balance within the adipose tissue. STAT3 activity is increased in visceral adipose tissue of obese mice and is also associated with increased IL-6 production, an inhibitor of Treg function. Ablation of STAT3 suppresses adipose tissue inflammation, increases the ratio of Treg/Th1 cells, and promoted M2 macrophage accumulation [
92]. It has recently been found that Tregs express the insulin receptor and that stimulation with high levels of insulin induces a decrease in their IL-10 production through activation of AKT signaling, thus contributing to obesity-associated inflammation. Moreover, the hyperinsulinemic mice fed with a high-fat diet showed a significant decrease in visceral adipose tissue Tregs IL-10 production and an increase in IFN-γ production [
93]. Lifestyle, nutritional, and pharmacological interventions aimed at restoring insulin sensitivity may also restore the Treg function in obese patients.
Interestingly, Poutahidis T. et al [
94] found an association between western diet-associated obesity, type of gut microflora, and CD4
+ Th17 prevalent T cell phenotype. Display of proinflammatory immune cell profile was prevented by microbial targeting that induced Foxp3
+ regulatory T cells and IL-10. Taken together, these findings support interventions aimed to enhance the anti-inflammatory properties of Tregs in humans and reduce the development of obesity-associated inflammation.
A novel subset of T helper cells, Th22 has been linked to chronic inflammatory conditions including obesity and diabetes. The proportion of circulating Th22 cells is increased in overweight/obese patients. Consistent with this observation, serum IL-22 level was significantly increased in obese patients when compared with lean subjects. The development of diabetes within the obese patient population led to further increase in circulating Th22 cells and IL-22 [
95] emphasizing the potential association between Th22 and the pathogenesis of obesity and type 2 diabetes.