Changes in adipose tissue macrophage and T cell during aging

Abstract

Adipose tissue historically was believed to be an inert tissue, functioning primarily in the storage of energy and thermal homeostasis. However, recent discoveries point toward a critical role for adipocytes in endocrine function as well as immune regulation. Excess body fat, accumulated through aging and/or calorie-rich diet, is associated with many chronic metabolic and inflammatory diseases. Within the stromal vascular fraction of adipose tissue, macrophages and T cells accumulate with increasing tissue mass, secreting pro- or anti-inflammatory cytokines. In this review we discuss the current understanding of immune cell function in both diet-induced and age-related obesity. In both models of obesity, the classically activated, pro-inflammatory (M1) subtype takes precedence over the alternatively activated, anti-inflammatory (M2) macrophages, causing tissue necrosis and releasing pro-inflammatory cytokines like IL-6. Recently, other distinct adipose tissue macrophage (ATM) subtypes have been identified by surface marker expression and their functions characterized. Adipose tissue T cell (ATT) recruitment to adipose tissue is also different between aging and diet-induced obesity. Under both conditions, T cells exhibit restricted T-cell receptor (TCR) diversity and produce higher levels of pro-inflammatory signals like IFN-γ and granzyme B relative to young or healthy mice. However, regulatory T cell numbers are dramatically different between the two models of obesity. Taken together, these findings suggest model of age- and diet-induced obesity may be more distinct than previously thought with many questions yet to be resolved in this multidimensional disease.

Introduction

Aging refers to the changes in a person’s physical, psychological and social ability with the passage of time. The aging mammalian immune system undergoes distinct and comprehensive changes (immune senescence), leading to progressive dysfunction of immune responses.¹ These changes occur at every level of the immune system and affect both innate and adaptive immune responses. Some aspects of the immune response intensify (such as inflammation, regulatory T cells frequency) while others are either unaffected or diminished (such as the decline in Th1-Th2 mediated immunity, less naïve T cell-mediated de novo responses, impaired dendritic cell function).^2–8 Age-related immune dysfunction is believed to be a key factor leading to increased susceptibility to infections, autoimmunity and cancer in elderly people.^9–11

Adipose tissue or body fat is a loose connective tissue composed mainly of adipocytes. It was earlier believed that the main function of fat was to store energy in the form of lipids as well as cushion and insulate the body. However, it has been shown that it is an active endocrine organ as well, producing a number of hormones such as leptin, estrogen, resistin, visfatin and cytokines such as tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6).^12,13 Through secretion of these factors, adipose tissue plays an important role in mammalian health and diseases, affecting other organ systems of the body through endocrine signaling.¹⁴ Based on its anatomical location, adipose tissue is classified as subcutaneous fat (under skin) or visceral fat (around internal organs). The two main types of adipose tissue in mammals are white adipose tissue (WAT) and brown adipose tissue (BAT). There are two distinct subtypes of BAT, depending on whether they are derived from myf-5 lineage and express the mitochondrial uncoupling protein UCP1. Additionally, investigators recently described CD137-high and UCP1-low expressing ‘beige’ adipocytes as a distinct type of thermogenic fat cell in mice and humans derived from progenitor cells within WAT.¹⁵ While BAT has been considered a specialized form of adipose tissue with the main role in generating heat by uncoupling the respiratory chain of oxidative phosphorylation, WAT has been associated with quality of life and lifespan in part through sirtuin 1 and peroxisome proliferator-activated receptor-γ (PPAR-γ).^16,17 Excess WAT is associated with cardiovascular disease, type 2 diabetes, insulin resistance and inflammatory diseases.^13,18–20

Macrophages and T cells are specialized leukocytes that protect the body against foreign invaders, repair tissue damage and clear dead cells from circulation. These cells perform their tasks partly by direct activity (such as phagocytosis in the case of macrophages) or by producing soluble factors (cytokines and chemokines). Macrophages and T cells work in a synergistic manner in order to mount an effective immune response. One of the adverse effects of the immune response is accumulation of inflammatory cytokines and chronic immune cell stimulation. Although acute healthy inflammation is beneficial, an unchecked response can lead to the development of chronic inflammatory diseases. Unfortunately, a state of low-grade chronic inflammation is commonly observed in the elderly. This phenomenon is known as “Inflamm-aging”, an immune condition characterized by the presence of greater amounts of pro-inflammatory cytokines compared to those found in young subjects.^1,21

Advancing age is increasingly accompanied by obesity, accumulation of visceral adipose tissue and inflammation.^22,23 The term obesity is not merely defined by excess weight but also depends on the accumulation of a large amount of fat. While the source of the inflammation in aging has not been definitively identified, the role of macrophages and T cells from adipose tissue has been shown to be associated with higher inflammation in aged mice.⁸ Lately, there is increasing interest in the field of age-induced obesity and inflammation; however, how age-induced obesity influences leukocyte-mediated inflammation is not clear. Given the attention paid to the intersection between aging, inflammation and adipose tissue, this current review highlights the changes in macrophage and T cell function during age-induced obesity and their implications for age-related diseases. Due to varying sources of cells (i.e. human or mouse), culture conditions and experimental protocols, it should be noted that there are inconsistencies in the literature and in some cases contradictory results.

Adipose tissue as an immune organ in obesity and aging

Adipose tissue is composed of several different cell types. The cellular composition of adipose tissue varies depending on locale and the metabolic health of the organism. Composition of the adipose tissue also depends on the lean state or physiological/pathological condition of the organism.^8,24,25 Physiologically, adipocytes are the major component of adipose tissue, but non-adipocytic cells, which make up the “stromal vascular cell fraction” (SVCF), also account for a significant proportion of adipose tissue. Separation of fat cells from the more dense SVCF has been described many years ago in pioneering protocol by Rodbell.²⁶ Using collagenase digestion and low speed centrifugation, adipocytes could be separated in the floating fraction from the dense SVCF pellet.²⁶ The dominant cell types in the SVCF are fibroblasts, macrophages, lymphocytes, pericytes and endothelial cells that are increasingly shown to perform a number of important functions for adipose tissue homeostasis. For example, monocytes and macrophages that reside in multiple fat depots, commonly known as adipose tissue macrophages (ATMs), help in the clearance of necrotic adipocytes, a role that appears to be of importance in obesity.²⁷ Several qualitative and quantitative phenotypic changes in ATMs that impact the inflammatory state of fat have been reported.⁸ The total percentage of SVCF consisting of ATMs increases during age-induced obesity.^8,28 Increase in the adipose tissue mass with aging is also due to differentiation of an existing pool of pre-adipocytes into adipocytes. Deferential gene and protein expression between pre-adipocytes and adipocytes revealed that major changes in adipose tissue are due to changes in pre-adipocytes which occur in parallel with an increase in inflammatory cytokines.^8,28,29

Hormones, cytokines, and chemokines produced in adipose tissue, collectively known as adipokines, play a prominent role in the attraction of immune cells to adipose tissue. Adipokines such as adiponectin, leptin, resistin, visfatin, TNF-α, IL-6, IL-1, plasminogen activator inhibitor 1 (PAI-1) and selected complement factors impact the inflammatory status of adipose tissue by affecting the quantity and action of immune cells that extravasate to the tissue.¹² Adiponectin, which is associated with lean states and insulin sensitivity, has been hailed as an anti-inflammatory force in adipose tissue via regulation of the production of cytokines in adipocytes and polarization of ATMs toward an M2 anti-inflammatory phenotype.^30,31 Leptin, produced in states of abundant adipose tissue and systemic inflammatory distress, induces the production of pro-inflammatory cytokines (such as TNF-α, IL-6, and IL-12) in monocytes.^32,33 Additionally, leptin promotes neutrophil chemotaxis and causes dysregulation of cell-mediated immunity, perhaps contributing to autoimmune diseases.³⁴ Resistin, made nominally in adipocytes and in greater quantities by circulating monocytes, increases expression of pro-inflammatory cytokines and endothelial cell adhesion molecules.³⁵ The pro-inflammatory adipocytokine profile (low adiponectin, and high leptin, TNF-a, and IL-6) observed in obese states has been hypothesized to be along a continuum of adipocyte dysregulation, stress, cell death, and immune system reactivation that is also seen in aging.³⁶ However, studies of healthy centenarians with low adipose tissue mass and high insulin sensitivity do not demonstrate elements of the pro-inflammatory profile, suggesting that adipocyte dysfunction may be a greater determinant of inflammation than aging.³⁶

Infiltration of macrophages in the adipose tissue is also shown to be responsible for obesity-associated inflammation and obesity-induced insulin resistance. For example, adipocytes secrete cytokines that act on endothelial cells to enhance vascular permeability, thereby allowing bone marrow-derived monocytes to migrate across vascular walls to localize to adipose tissue and differentiate into ATMs.³⁷ Although the complete spectrum of factors attracting the monocytes/macrophages is not fully understood, leptin in adipocyte-conditioned media may upregulate adhesion molecules on endothelial cells, permitting increased chemotaxis of blood monocytes.^37,38 Myeloid-specific deletions of inhibitor of nuclear factor kappa-b (IKKb), CC chemokine receptor-2 (CCR2), toll like receptor-4 (TLR4), Cbl-asscociated protein, fatty acid binding protein 4, or depletion of CD11c+ cells in mice have been shown to ameliorate obesity-induced inflammation, raise glucose tolerance and lower insulin resistance.^39–44 Moreover, germline deletion of Ccr2 gene resulted in reduced macrophage infiltration, adipose tissue inflammation, and improved insulin sensitivity in a mouse model fed a high fat diet (HFD).⁴⁵ Other studies have shown that PPAR-γ and PPAR-δ serve a protective role in immune and metabolic homeostasis, as deletion of these genes worsen inflammation and switch the native M2 (anti-inflammatory) macrophage into M1 (pro-inflammatory) phenotype.^46,47 In addition to macrophages, the number of mast cells and natural killer T (NKT) cells also increase in obese adipose tissue compared with lean tissue and may contribute to the inflammatory and metabolic pathophysiology.^48,49

Adipose tissue T cells (ATT) contribute to the pro- inflammatory environment in visceral fat during normal aging.⁸ The role of ATT-cells in obesity-induced inflammation has also been reported. Recent works report altered T cell homeostasis in obesity, likely due to a decrease in Th2 cells and an increase in CD8+ T cells.^50–52 Moreover, the number of regulatory T cells (CD4⁺CD25⁺, Tregs) decrease as a function of obesity, which may contribute to a hyper immune activation. In fact, depletion of CD8⁺ cells, NKT, or mast cells or enhancement of CD4⁺ cells or Treg cells decreased macrophage recruitment and inflammation in the adipose tissue and improved glucose homeostasis.^48–52 These data clearly show that adipose tissue-resident immune cells are important contributors to obesity-induced inflammation and metabolic syndrome. Thus, adipose tissue can be thought of as an immune organ in the context of both aging and obesity.

Macrophage subtypes in adipose tissue

Macrophages are highly heterogeneous hematopoietic cells produced by the differentiation of monocytes in tissue. They are specialized phagocytic cells that clear foreign substances, infectious microbes, and cancer cells by ingestion and destruction. In addition to their role in innate immunity, macrophages are important sentinels of the adaptive immune response, wound healing, and tissue repair. Not surprisingly, there is no uniformly accepted classification of the many macrophage subtypes. Mouse macrophages can be identified by the expression of several surface markers such as CD14, CD40, CD11b, F4/80, and CD68. The human ortholog of F4/80 is EMR1. Historically, macrophages were classified under the prototypical dichotomy of M1 “classically” activated macrophages and M2 “alternatively” activated macrophages (Table 1).^53,54 Undifferentiated macrophages exposed to lipopolysaccharide (LPS, a cell wall component of gram negative bacteria) or to interferon-γ (IFN-γ) give rise to M1 macrophages (CD11c⁺CD206⁻). These cells have high phagocytic and bactericidal potential, secrete pro-inflammatory cytokines and activate Th1 lymphocytes. In contrast, alternative activation occurring in the presence of IL-4, IL-13, or parasitic infection generates M2 macrophages (CD11c⁻CD206⁺). These macrophages interact with Th2 lymphocytes to promote anti-parasitic activity, wound healing and tissue repair as well as produce anti-inflammatory cytokines (e.g. IL-10) that prevent excessive immune responses (Table 1).^7,55 The M2 population is further subdivided into M2a (initiates type II inflammation and fibrosis), M2b (immunoregulation/immunosuppression), M2c (participate in matrix remodeling and tissue repair), and M2d (tumor-associated).^56,57 It has been proposed that some macrophages may have regulatory functions.⁵⁸ However, a recent report of Foxp3-expressing regulatory macrophages (Mregs) that secrete large amounts of PGE2 was retracted.⁵⁹

Table 1.

Macrophage (CD11b⁺F4/80⁺) subtypes in mice.

	M1 (classically activated macrophages)	M2 (alternatively activated macrophages)
Marker	CD11c+CD206−, MHC-IIhigh, CD86	CD11c−CD206+high
Stimuli	IFN-γ, LPS, TNF-α	IL-4, IL-13, IL-10, TGF-β
Secrete/Produce	ROS, RNS, TNF-α, IL-1, IL-6, IL-12, IL-23	IL-10, TGF-β, PDGF, VEGF, EGF, Arg-1
Function	help Th1 cells, anti-microbial, inflammation, cytotoxicity, tissue injury, phagocytosis, tumor suppression	help Th2 cells, anti-parasitic (all M2). type II inflammation and fibrosis (M2a); anti-inflammation and immune suppression (M2b); matrix remodeling, wound healing and tissue repair (M2c); tumor promoting (M2d).
Reference	[53, 54]	[7, 8, 53, 54, 56]

CD: cluster of differentiation, MHC: major histocompatibility complex, IFN: interferon, TNF: tumor necrosis factor, IL: interleukin, TGF: transforming growth factor, LPS: lipopolysaccharide, ROS: reactive oxygen species, RNS: reactive nitrogen species, PDGF: platelet-derived growth factor, VEGF: vascular endothelial growth factor, EGF: epidermal growth factor, Th: helper T cell,

Adipose tissue macrophages (ATMs; CD11b⁺F4/80⁺) are the greatest proportion of leukocytes in adipose tissue. While the M1/M2 paradigm has been a useful and straightforward classification of tissue macrophages in previous studies, ATMs exist in a spectrum of inflammatory and activation states, thus requiring an expanded definition of ATM classification. One proposed classification of ATMs in mice by Morris et al renames M1 ATMs as Type 1a ATMs (CD11c⁺CD206⁻ MGL1⁻), maintains M2 ATMs as Type 2 ATMs (CD11c⁻CD206⁺ MGL1^hi), and introduces the two new subtypes Type 1b ATMs (CD11c⁺CD206⁺ MGL1^mid) and Type 3 ATMs (CD11c⁻CD206⁻) (Table 2). Morris et al also introduces a similar classification of ATM subtypes for humans as Type 1a ATMs (CD11c⁺ CD206⁻), Type 1b ATMs (CD11c⁺ CD206⁺), and Type 2 ATMs (CD11c⁻ CD206⁺). However, the double negative Type 3 ATM phenotype has not been observed in humans.⁶⁰ Lumeng et al proposed a similar classification of ATMs in mice that maintains the M1 and M2 ATM subtypes and names the inflammatory double-negative macrophage (CD11c⁻CD206⁻) as a type 4 ATM.⁸

Table 2.

Adipose tissue macrophage (ATMs; CD11b+F4/80+) subtypes in mice.

	Type 1a ATMs	Type 1b ATMs	Type 2 ATMs	Type 3 ATMs
Marker	CD11c+CD206− MGL1−	CD11c+CD206+MGL1mid	CD11c−CD206+ MGL1hi	CD11c− CD206−
Gene expression	IL-6, TNF-α, Nos2, Ccr-7	IL-12, Arg-1	IL-10, Arg-1, Ccl-2, Mgl1/2, YM1/Chi3/3	Arg-1, Chi3/3, Ccl-2, Ccr-1, Ccr-9
Change with obesity/lean	Increase	Increase	decrease or no	Increase
Change with aging/young	no change or increase	no change or increase	decrease	increase
Localization	CLS	CLS	Interstitial	CLS
Function	Inflammatory	Inflammatory	Anti-inflammatory	Inflammatory
Resemblance	classical M1 type	classical M1 type	classical M2 type	Mix of M1 and M2
Reference	[8,60,61,62]	[8,60,61,62]	[8,60,61,64]	[8,60,61,62]

CD: cluster of differentiation, TNF: tumor necrosis factor, IL: interleukin, Arg: arginase, CCr: CC chemokine receptor, MMGL: macrophage galactose N-acetyl-galactosamine-specific lectins, YM1/Chi3/3: chitinase 3-like 3 protein, CLS: crown-like structure.

It is generally accepted that inflammatory M1 ATMs are more prevalent as a percentage of total macrophages in the adipose tissue of obese animals,⁸ while protective M2 ATMs are strongly associated with the adipose tissue of lean animals (Table 3, 4). Similarly, M2 ATMs seem to decrease and DN Type 3 ATMs seem to increase in quantity with increasing age.⁸ However, transcriptional profiles of ATM subtypes did not always reveal the expected subset associated prototypical markers and cytokines. Type 1a and Type 1b ATMs isolated in obese mice by Shaul et al expressed a mixed profile of M1 and M2 traits.⁶¹ The CD11b⁺ ATMs had upregulated pro-inflammatory genes such as IL1β and Th1-priming IL12p40 as well as M2-profile tissue homeostasis protein arginase 1 (Arg1) and IL1Ra (IL-1 receptor antagonist).⁶¹ All ATMs identified in this study, which included Type 1a, Type 1b, and Type 2 ATMs, had a tendency toward M2 anti-inflammatory and tissue homeostatic traits. Similarly, Zeyda and coworkers also identified Type 1, Type 2, and Type 3 ATMs in diet-induced obese mice but Type 2 ATMs isolated in this study did not exhibit the typical anti-inflammatory profile of alternatively activated M2 macrophages.⁶² Rather, Type 2 ATMs expressed the most chemokines (Ccl2, Ccl5, Ccl8, Ccl11, and Cx3cl1) out of the three types of ATMs, indicating they may play a significant role attracting immune cells to adipose tissue during obesity and infection, rather than attenuating inflammatory responses. Interestingly, Type 3 ATMs demonstrated M2 ATM traits of anti-inflammatory action and tissue repair based on its expression of Ym1/Chi313 (YM1/chitinase 3-like 3) and Arg1 to a greater extent than Type 2 ATMs, which minimally expressed those proteins in this study. These observations of complex properties of Type 1a, Type 1b, Type 2, and Type 3 ATMs suggest a phenotypic plasticity of macrophages, expose the limitations of the previous M1/M2 dichotomy label of ATMs, and demonstrate the need for further investigation to more clearly elucidate subtype characteristics and functions.

Table 3.

ATMs in obese versus lean mice

	ATMs from obese mice	ATMs from lean mice
Markers	CD11c+CD206−	CD11c−CD206+high, MGL1+high
Expression	IL-1β, IL-6, Nos2, TNF-α	Arg-1, Ym-1
Function	Inflammatory	Anti-inflammatory, insulin sensitivity
Localization	CLS	interstitial

Table 4.

Changes in ATMs proteins expression due to either obesity or aging in respect to lean or young state respectively.

	Obesity/Lean	Old/Young
IL-6	increase [65,66,67]	increase [8]
MCP-1	increase [66]	increase [8]
TNF-α	increase [64,66,68]	increase [8]
Ccr-2	increase [45,62,69]	increase [8]
Cx3cr-1	no change or increase [62,66,70]	decrease [8]
Ccr-5	increase [62,69]	increase [8]
Ccr-7	increase [62]	decrease [8]
Cxcr-3	nd	increase [8]
PPAR-γ	increase [66]	decrease [77]
NOS2	increase [62,64]	nd
ApoE	decrease [64]	nd
Ym1/Chi3/3	decrease [62,64]	decrease
IL-10	decrease [64]	no change [8]
IL-1β	Increase [60]	increase [77]

IL: interleukin, MCP: monocyte chemoattractant protein, TNF: tumor necrosis factor, Ccr: CC chemokine receptor, Cxcr: CXC chemokine receptor, PPAR: peroxisome proliferators-activated receptor, ApoE: apoliporotein E, YM1/chi3/3: chitinase 3-like 3 protein, nd: not determined.

Role of ATMs in obesity associated inflammation

ATMs infiltrate visceral adipose tissue in greater numbers than subcutaneous adipose tissue during obesity.⁶³ In both lean and obese states in mice, Type 1a, Type 1b, and Type 3 ATMs localize to adipose tissue as crown-like structures (CLS) of up to 15 macrophages surrounding one adipocyte.^27,60 However, CLS are rare in the adipose tissue of lean animals as pro-inflammatory ATMs do not infiltrate adipose tissue in significant numbers until the onset of obesity or pathology in which adipocytes are more prone to necrosis.⁶² CLS have been observed to increase 30-fold in obese mice and humans, accounting for more than 90% of the ATMs in adipose tissue. If adipocyte death is persistent, as exhibited in obesity, the ATMs of CLS can recruit more macrophages and develop into multinucleated giant cells that augment the chronic inflammation seen in obesity.

ATMs are major contributors to the chronic pro-inflammatory state of fat in obese individuals.²⁸ Feeding mice a HFD increases obesity and shifts the ATM profile from anti-inflammatory, protective Type 2 ATMs to pro-inflammatory, destructive Type 1 ATMs.⁶⁴ During obesity, the inflammatory genes that Type 1 ATMs express are IL-6 and Nos2, IL-1β, Ccr2, Cx3cr1, Ccr5, Ccr7 PPAR-γ; Type 1 and Type 2 ATMs express equal amounts of TNF-α; and Type 2 ATMs express more ApoE.^{45,62,64–70} Lean mice ATMs are predominantly the Type 2 variety, and they express the characteristic anti-inflammatory genes IL-10, Arg1, Mrc2 (mannose receptor C type 2, also known as CD206), Ym1/Chi313 (YM1/chitinase 3-like 3), and Mgl1/2 (macrophage galactose N-acetyl-galactosamine-specific lectins 1 and 2) (Table 3, 4).^62,64 Greater numbers of Type 1 ATMs and loss of Type 2 ATMs in obesity result in the loss of the anti-inflammatory, protective functions of Type 2 ATMs in maintaining adipose tissue homeostasis and repair. Rather, Type 1 ATMs have carte blanche to secrete pro-inflammatory cytokines and perpetuate the inflammatory state.

Role of ATMs in Inflamm-aging

Natural selection favors a strong immune system that protects humans well into their reproductive years, but the dramatic increase in lifespan in recent centuries has exposed negative consequences of a persistent immune response in the elderly. Thus, although an acute inflammatory response to infection and injury is critical to life, the chronic low grade inflammatory state seen in the elderly increases the individual’s susceptibility to age-related diseases. In aging, high levels of inflammatory cytokines such as IL-3, IL-6, IL-8, IL-10, IL-15, and TNF-α are coupled with a decrease in naïve T cell compartment, leading to chronic low-level inflammation yet decreased ability to fight off novel infections.⁷¹ NK cells increase in number, and macrophages demonstrate increased secretion of pro-inflammatory signals like prostaglandins.⁷² The negative consequences of overactive innate immunity, chronic inflammation, and T cell senescence are referred to as ‘inflamm-aging.’

While the source of the low grade elevated inflammatory cytokines in aging has not been defined, it is believed that fat tissue inflammation in aging-associated obesity may play a role. Diverse types of leukocytes are present in fat tissue and act as cellular mediators of inflamm-aging and contribute to insulin resistance. Normal aging often leads to an increase in visceral and subcutaneous adiposity. This buildup of adipose tissue has been implicated as the source for elevated inflammatory cytokines observed in obese individuals.^73,74 Indeed, it has been suggested that 30% of circulating IL-6 may be derived from adipose tissue.⁶⁷ Comparing lean old horses with fat old horses demonstrated that increased adiposity led to greater numbers of T cells and monocytes producing inflammatory cytokines IFN-γ and TNF-α, a phenotype which could be mitigated by feeding the obese animals a calorie-restricted diet for several weeks.⁶⁸ Therefore, normal aging leads to a more pro-inflammatory profile,^75,76 further accentuated by increasing adiposity. The M1/M2 paradigm of macrophage polarization explains the phenotypic switch to this inflammatory pathway. Concomitant with greater body fat percentage, aged mice also have a greater number of ATTs and ATMs than young mice.⁸ These macrophages were also more likely to possess a pro-inflammatory phenotype than their younger counterparts, as aging decreases M2 while increasing the pool of M1 and double negative (DN) macrophage subsets, and they produce greater quantities of TNF-α, IL-6, IL-1β and monocyte chemoattracting protein-1 (MCP-1), Ccr2, Ccr5, Cxcr3, Ym1/Chi3/3.^8,77 In contrast to obesity versus lean model, aging is manifested with decrease in Cx3cr1, Ccr7 and PPAR-γ expression in respect to young mice (Table 4).^8,77 The contribution of adipose tissue to inflamm-aging is supported by numerous studies demonstrating aged individuals who exercise regularly and are leaner have less senescent T cells and lower circulating pro-inflammatory cytokines.⁷⁸ Acute exercise also can repolarize ATMs toward the anti-inflammatory M2 phenotype in rats fed high fat diet.⁷⁹ Regular exercise can maintain skeletal muscle and prevent accumulation of visceral fat, resulting in a more ‘youthful’ body composition.⁸⁰ Therefore, higher levels of circulating inflammatory signals derived from greater body fat percentage in aged individuals suggest adipose tissue-resident immune cells may be key contributors toward the sub-clinical phenomenon of inflamm-aging.

Contribution of T cells subtypes in age- and obesity-related inflammation

T cell homeostasis (number, subset ratio, function and phenotype) changes during normal aging and obesity. There are several subtypes of T cells that play a role in cell-mediated immunity- T helper cells or CD4⁺ cells; cytotoxic T cells or CD8⁺ cells; regulatory T cell or CD4⁺CD25⁺ cells; memory T cells; NKT cells; and γδ T cells- each with distinct number, phenotype and function under physiological conditions. A vast array of literature suggests that aged spleens and lymph nodes, both from humans and mice, show an increase in the proportion of memory T cells and a reciprocal decrease in naïve T cells.^81–83 A longitudinal study from genetically heterogeneous mouse peripheral blood leukocytes (PBLs) reports that aging led to an increase in the total number of CD3⁺ cells, CD4⁺ and CD8⁺ memory T cells and declines in the proportion of CD4⁺ and CD8⁺ naïve T cells.⁸⁴ While changes in the T cell repertoire occur during normal aging, changes in the chemokine and cytokine profiles contribute to a pro-inflammatory environment.^85,86 Freshly isolated CD4⁺ cells from aged mice were shown to express a higher level of several T cell chemokine receptors (such as CCR1, 2, 4, 5, 6, and 8; CXCR2,3,4 and 5) and lower level of CCR7 and CCR9 in relation to young mice.⁸⁷ T cells from old mice also showed higher chemotaxis to stromal cell-derived factor-1 and monocyte inhibitory protein-1α (MIP-1α). Repeated observations have shown that T cells from healthy elderly people secrete higher level of several cytokines including TNF-α, IL-6 and C-reactive protein (CRP) relative to their younger counterparts^6,88 that correlate with increasing overall mortality and/or frailty.⁸⁹

While the underlying mechanism for the observed aging-associated changes in immune function is unclear, a shift in memory/naïve profile and Th1/Th2 ratio may account partly for some of the aging-associated changes in immune function. However, based on the recent work on changes in chemokine and cytokine systems in aged individuals, other mechanisms may also contribute to the aging phenotype. For example, the Ccr2 gene has a binding site for C/EBP, a transcription factor that interacts with p50 subunit of nuclear factor kappa-B (NF-kB) and activates CCR5 expression.^90–92 NF-kB is also shown to be an important regulator of a number of chemokine receptor genes.^93–95 Since NF-kB activity is shown to diminish with age, it is unlikely that observed age-associated changes in the chemokine system are entirely due to changes in the transcriptional activity.^96,97 Maintenance of epigenetic marks declines with age;⁹⁸ thus, it is possible that epigenetic alteration could play a crucial role in greater inflammation in the elderly. Several chemokine/cytokine pathways implicated in obesity-induced inflammation have been shown to be regulated by epigenetic modulation.^99,100 DNA methylation can regulate gene expression at proximal promoters, upstream and downstream enhancers, and can alter chromatin structures which may regulate gene expression over large distances. Our lab has shown previously that Ccr2, Cxcr3, Ccr5 messages and TNF-α, IFN-γ, IL-4 and IL-2 cytokines are sensitive to global epigenetic perturbation. The levels of these chemokines and cytokines were significantly lower in T cells exhibiting hypermethylated DNA derived from mice born from dams fed with a methyl-donor rich diet. This intervention also protects against atherosclerosis in ApoE^−/− mice fed a high fat diet.¹⁰¹ In addition, histone modification also changes the higher order structure of chromatin, often working in tandem with DNA methylation patterning.

Given that ATMs and ATTs work synergistically and ATM are responsible for inflammation in the adipose tissue and T cell-derived cytokines are required for migration and activation of macrophages, it is plausible to ascribe to T cells a role in adipose tissue inflammation and insulin resistance.¹⁰² It has been shown that diet-induced obesity activates ATTs that in turn alter ATM phenotype. Moreover, it has been shown that changes in ATMs are parallel with the similar changes in the ATTs.⁸ In addition, Lumeng et al have demonstrated that aging is associated with a 2 fold induction of ATT cells in visceral fat tissue but not in the spleen, suggesting that this change is unique to adipose tissue (Table 5).⁸ While studying the changes in ATT subpopulations, the investigators observed a significant increase in the percentage of both CD4⁺ and CD8⁺ cells in adipose tissue in old mice. They also report that the expansion of CD4⁺ T cells in the aged adipose tissue parallels an increase in the size of fat-associated lymphoid clusters (FALCs). In the case of diet-induced obesity, it was demonstrated that total CD3⁺, CD4⁺ and CD8⁺ T cells were dramatically increased in the epididymal tissue (Table 5).¹⁰³ Furthermore, obesity reduced the frequency of adipose-resident naïve T cells in the subcutaneous fat and increased the effector memory population in the visceral fat. ATTs from obese mice produce higher pro-inflammatory mediators such as IFN-γ and granzyme B relative to lean mice. Similar to peripheral T cells from aged mice, adipose resident T cells also exhibit restricted TCR diversity, which is further compromised by obesity.⁸

Table 5.

Quantitative changes in T cell populations as a function of aging and obesity

	Naïve CD3+	Naïve CD4+	Naïve CD8+	nCD4+CD25+
Aged SVF [8]	increase	no change or increase	no change or increase	increase
Aged spleen [8]	decrease	decrease	decrease	no change
Obese SVF [103]	increase	increase	increase	decrease
Obese spleen [103]	nd	decrease	nd	increase

SVF: stromal vascular fraction, nCD4⁺Cd25⁺: naturally occurring CD4⁺CD25⁺, nd: not determined.

Regulatory T cells (Treg) which constitute about 5–20% of the CD4⁺ T cells are thought to maintain peripheral tolerance and keep autoreactive cells at bay. Treg can be defined by the expression of the forkhead/winged-helix family of transcription factor (Foxp3) that is both necessary and sufficient for Treg development.¹⁰⁴ Two major populations of Tregs exist: so-called natural (nTreg) and inducible (iTreg) Tregs. nTregs mature in the thymus and iTreg are induced extrathymically in the secondary lymphoid organs in response to antigen exposure.¹⁰⁵ Normally, Tregs control the activity of other T cell populations, maintain lymphoid homeostasis and prevent auto-reactivity. However, excess Treg activity may lead to increased susceptibility to infection, neurodegenerative diseases and cancer.^9–11 This is particularly important during normal aging where increased frequency of Treg might contribute to immune senescence.

Tregs have also been found in adipose tissue. It is believed that Treg cells in the fat provide anti-inflammatory signals to block adipose tissue inflammation. While comparing fat from lean and obese animals, several studies revealed that the latter adipose tissue had decreased numbers of Tregs, perhaps contributing to higher levels of inflammation and decreased insulin sensitivity (Table 5).^50,52,103 Boosting Treg number increased the insulin sensitivity partly through secretion of IL-10 which also controls ATM-mediated inflammation.⁵⁰ TNF-α and IL-6 are higher in circulation in the obese mice. TNF-α signal adversely affects adipocyte and decreases insulin sensitivity. Feurer et al also show that Treg-mediated IL-10 not only blocks adipocyte-derived inflammatory mediators but also protects adipocytes from the negative effect on insulin signaling induced by TNF-α.⁵⁰ Moreover, recent work done by our group showed an 11-fold induction in adipose tissue Treg number as a function of aging (Table 5).⁸ These data clearly suggest that obesity decreases the number of Treg frequency in the adipose tissue while normal aging increases it. The difference in adipose tissue Treg frequency in aging and diet-induced model of obesity could be attributable to either a differential inflammatory or nutrient metabolism environment. However, our understanding about an underlying mechanism that promotes an increase in adipose tissue Treg in one case versus decrease in the other is incomplete and requires further investigation.

Summary

The role of adipose tissue in regulating systemic inflammation is only now beginning to be elucidated. Novel macrophage cell types and sites of lymphoid localization have been described in the fat. Within the context of obesity, whether diet-induced or aging-related, excess body fat creates a pro-inflammatory environment, sustaining chronic low level inflammation and decreased immune cell plasticity, the hallmarks of inflamm-aging. While macrophages seem to behave similarly in the two models of obesity, T cell subsets are different, especially the Treg compartment. Since T cells are responsible for chemotactic signals that recruit monocytes to the fat to differentiate into ATMs, the observed difference suggests the aging model of obesity may not be as similar to the diet-induced model as is currently understood. More investigation into the role of adipose tissue function is needed to clarify the phenotype, function and number of macrophage subsets, the role of T cell subsets in mediating signals that recruit immune cells, and the relationship between diet and aging in the development of obesity-induced systemic inflammation.

Abbreviations

TCR

T cell receptor

ATM

adipose tissue macrophage

ATT

adipose tissue T cell

SVCF

stomal vascular cell fraction

SVF

stromal vascular fraction

HFD

high fat diet

CD

cluster of differentiation

MHC

major histocompatibility complex

IFN

interferon

TNF

tumor necrosis factor

IL

interleukin

LPS

lipopolysaccharide

ROS

reactive oxygen species

RNS

reactive nitrogen species

PDGF

platelet-derived growth factor

VEGF

vascular endothelial growth factor

EGF

epidermal growth factor

Th

helper T cell

NKT

natural killer T cell

Tregs

regulatory T cell

nTreg

naturally occurring regulatory T cells

Nos

nitric oxide synthetase

TGF

transforming growth factor

Arg

arginase

Ccr

CC chemokine receptor

Cxcr

CXC chemokine receptor

MMGL

macrophage galactose N-acetyl galactosamine-specific lectins

YM1/Chi3/3

chitinase 3-like 3 protein

CLS

crown-like structure

MCP

monocyte chemoattractant protein

PPAR

peroxisome proliferator-activated receptor

ApoE

apoliporotein E

FALCs

fat-associated lymphoid clusters