Introduction

Obesity, i.e. the abnormal or excessive accumulation of body fat that presents a risk to health wherein the body mass index (BMI) is > 30, is a condition with increasing prevalence, which affected 10.8% of men and 14.9% of women worldwide in 2014 [1]. Obesity accounted for 9% of the cancer burden in women in Europe, America, and the Middle East. The risk of cancers in obese individuals is increased, with relative risks (RRs) ranging from 1.5 to 1.8 for cancers of the colon, gastric cardia, liver, gallbladder, pancreas, and kidney [1–3]. Moreover, the RR for esophageal adenocarcinoma was up to 4.8 for a BMI of 40 or more [1].

Effects of obesity on the functions of the immune system are thought to be an important mechanism responsible for the increased risk of cancer. A review of pre-clinical murine studies by Farag et al. [4] showed that obesity induces lymph node atrophy, impedes lymphatic transport, and reduces T cell receptor (TCR) diversity. Other effects included accumulation of tumor-promoting myeloid cells, e.g., macrophages, neutrophils, and myeloid-derived suppressor cells (MDSCs) within the tumor microenvironment [4]. In humans, the function of type 1 innate lymphoid cells (ILC-1) in adipose tissue (AT) that regulate macrophages via cytotoxicity is impaired in obesity, thus contributing to a pro-inflammatory phenotype [5]. Obesity impairs the function of anti-cancer natural killer (NK) cells by impeding their ability to direct lytic granules to the tumor cell and degranulate [6]. Interestingly, impaired NK activity characteristic of obesity can be improved by weight loss [7].

γδ T cells, i.e. T cells expressing TCR encoded by the γ and δ genes, were discovered in the 1980s, and constitute one of the three arms of the immune system (the other two being αβ T cells and B cells) characterized by surface membrane expression of somatically rearranging immunoreceptors [8–10]. γδ T cells emerged evolutionarily about 450 million years ago, along with B cells and αβ T cells, and remain the least well understood [11]. Some outstanding unique features of γδ T cells include their characteristic lack (on the majority of cells) of expression of cluster of differentiation 4 (CD4) and CD8 (unlike αβ T cells), their early establishment of residence in surface epithelial tissues, and their independence from major histocompatibility complex (MHC) molecules for antigen recognition [12–14]. Indeed, the manner wherein the γδ TCR recognizes antigens is unique although differing between subsets of γδ T according to the variable (V) regions of the TCR. Thus, the majority of γδ T cells expressing Vγ9 and Vδ2 in the TCR (Vγ9δ2 cells), which form the major subset in human peripheral blood (PB), innately recognize cell surface membrane expressed butyrophilin (BTN) 3A1 and 2A1 molecules when they assume a novel conformation induced by intracellular binding of phosphoantigens, upregulation of which occurs during oncogenic transformation or intracellular infections [15]. On the other hand, Vδ1⁺ γδ T cells, which are chiefly located in tissues along epithelial surfaces, recognize MHC-like molecules including CD1d, a, b, and c, either directly, or when presenting lipid antigens including sulfatides [15]. Furthermore, Vδ1⁺ or Vδ3⁺ clones express TCR that directly recognizes sites on cell surface membrane proteins including MHC class I-related 1 (MR1), endothelial receptor C protein, annexin, phycoerythrin, and ephrin A. γδ T cells also express NK like receptors, e.g., NK receptor group 2 member D (NKG2D) that can mediate cytotoxicity by interactions with MHC class I polypeptide-related sequence A (MICA) or MICB [15]. Thus, γδ T cells appear to exhibit a functional repertoire straddling innate and adaptive immunity [13, 14, 16].

The very first paper describing the discovery of human thymus-derived T cells expressing a γδ TCR already described the cytotoxic potential of γδ T cells and the potential role of the CD3–TCR complex in mediating killing against hematological tumor cells [8]. This anti-tumoral cytotoxic potential has been confirmed by numerous studies suggesting that these cells may play a strong anti-tumoral role in humans further stimulated by the revelation that, relative to other leukocytes and immunocytes, the presence of γδ T cells in the tumor stroma of a large cohort of cancer patients was the best positive prognostic indicator [17, 18]. The mechanisms include Vγ9δ2 TCR interactions with BTNs bound to intracellular tumor phospho-antigens, as well as recognition of stress-related molecules, expressed on tumor cell membranes, which result in perforin and granzyme B mediated killing of tumor cells [17]. In addition, NK cell-like NKG2D mediated killing by recognition of MICA and MICB has been implicated [17]. Cytotoxic molecules such as tumor necrosis factor α (TNFα) may act directly on tumor cells and type I cytokines interferon γ (IFNγ) that enhance the anti-tumoral response of adaptive immune cells, both of which are highly secreted by γδ T cells, are additional important mechanisms of anti-tumoral effects of γδ T cells [17]. Finally, activated Vγ9δ2 cells may function as tumor antigen-presenting cells, activating tumor antigen-specific CD8⁺ and CD4⁺ αβ T cells [19].

Despite these potent anti-tumoral mechanisms, there are strong indications that γδ T cells may also critically play pro-tumoral roles [20]. For example, immunosuppressive γδ T cells have been shown in breast cancer and pancreatic cancer [21–23]. Another important mechanism involves γδ T cells that secrete interleukin-17 (IL-17). Thus, IL-17 secreting T cells were shown to enhance breast cancer growth and metastasis in mice [24]. In pancreatic cancer, IL-17 producing γδ T cells are a poor prognostic factor and were shown in murine models to directly enhance tumor growth [25, 26]. In murine models of lung cancer, IL-17 producing γδ T cells triggered by the microbiota were shown to mediate cancer growth [27]. The mechanism may involve IL-17 mediated promotion of pro-tumoral neutrophils, promotion of vascular endothelial growth factor (VEGF) to enhance tumor angiogenesis, or enhancement of pro-tumoral mesenchymal stem cells [3, 28, 29].

Despite growing interest in the role of γδ T cells in cancer, there are no published reviews of how obesity impacts this important arm of the immune system. Critically, although there exists considerable functional overlap (e.g., cytokine repertoire, cytotoxic, helper and regulatory) of γδ T cells and other T cells of either “adaptive” (classical MHC restricted CD4⁺ and CD8⁺ αβ T cells) or of “innate” type [including NKT and mucosal associated invariant T (MAIT) cells] their highly unique antigenic repertoire, which includes phospho-antigen metabolites in the cholesterol synthetic pathway, as well as lipid antigens, strongly suggests that they could play a pivotal role in the effect of obesity on cancer [13, 14, 16]. Here, we will present the available evidence addressing the impact of obesity on the distribution and functions of γδ T cells in animal models and in humans, and evaluate how obesity-induced changes might influence the immune effects of γδ T cell in the setting of cancer.

Physiology of γδ T cells in AT

γδ T cells, similarly to other classical and innate like T cells, are present in AT in mice, humans, and cows. In cows, the frequency of both γδ T cells and CD8⁺ non-γδ T cells was found to be higher in mesenteric than in subcutaneous AT and expressed a CD45RO⁺CD62L^– phenotype [30]. Importantly, in murine models, it was found that the predominant cells secreting IL-17 in AT are γδ T cells. Moreover, IL-17 produced by these cells was shown to inhibit adipogenesis, to moderate AT accumulation, and to regulate glucose metabolism [31]. Adipose T cells promyelocytic leukemia zinc finger (PLZF)⁺ γδ T cells producing both TNFα and IL-17 induced IL-33 production by platelet-derived growth factor receptor α (PDGFRα) expressing stromal cells, which by its effects on suppression of tumorigenicity 2 (ST2)⁺ regulatory T (Treg) cells controlled non-shivering thermogenesis. Thus, γδ knock out mice exhibited loss of temperature regulation, in particular in cold conditions [32]. The IL-17 produced mainly by γδ T cells in AT binds to the IL-17 receptor C (IL-17RC). Ablation of IL-17RC reduced transforming growth factor β1 (TGFβ1) expression which in turn reduced sympathetic innervation in AT, induced obesity and reduced thermogenicity. Accordingly, reduction of IL-17 producing γδ T cells had similar effects. Interestingly, ablation of IL-17RC or γδ T cells also reduced sympathetic innervation in the lung and salivary glands [33]. Others confirmed that mice lacking γδ T cells have impaired glucose homeostasis. Interestingly, a ketogenic diet (KD) expanded metabolically protective γδ T cells that restrain inflammation. However, long term ad libitum KD induces obesity and impairs metabolic health while depleting AT fat tissue γδ T cells, suggesting that γδ T cells mediate protective immunometabolic response and thus decrease AT inflammation [34]. On the other hand, γδ TCR expressing intraepithelial lymphocytes (IEL) in the gut, may have an opposite effect in terms of immunometabolic control. Thus, by expressing receptors for glucagon-like peptide-1 (GLP-1), γδ, as well as αβ IEL, serve as a “sink” for this important metabolic regulator by decreasing its bioavailability, thus contributing to insulin resistance (IR) and obesity. Indeed, mice lacking IEL by virtue of ablation of α4β7 and αeβ7 integrins, are resistant to the effects of a high fat diet (HFD) [35].

On the other hand, previously, it had been shown that an HFD which induced obesity in mice, also increased γδ cells in inguinal fat [36] and that the γδ T cells increase in fatty tissue in obese mice was less pronounced in leptin-deficient mice, suggesting a role for leptin in the control of AT γδ T cells [37]. An additional factor controlling the homeostasis of γδ T cells in obesity may be TNF-like cytokine 1A (TL1A), a TNFα-related molecule [38]. Thus, TL1A deficient mice were both less obese and had fewer γδ T cells in epididymal white fat, concomitant with a reduction of mitochondria regulatory genes. Interestingly, in humans, a non-coding polymorphism close to the TL1A locus was associated with body mass, suggesting that studies of AT γδ cells in these individuals would be of interest [38]. Others have reported that HFD may specifically induce an increase of CD44⁺CD62lo γδ T cells in epididymal fat of mice. In this study, although Vγ4 and 6 deficient mice or TCRδ^–/– mice were no different in weight relative to wild type (WT), they had less inflammatory macrophages producing IL-6 in their AT—suggesting a proinflammatory role of Vγ4 and 6 expressing γδ T cells, which are the main IL-17 producing γδ T cells, in visceral AT (VAT). Interestingly, however, in this study, TCRδ^–/– mice were less insulin resistant and had reduced levels of inflammation in the liver and skeletal muscle [39]. In another study, it was shown that decreasing αβ/γδ T cell ratio in white AT (WAT) depots alters their inflammatory state and mass repartition, which might be involved in the improvement of insulin sensitivity. Thus, a mouse model overexpressing peroxisome proliferator-activated receptor β (PPAR-β) specifically in T cells [transgenic (Tg) T-PPAR-β] exhibits a partial depletion in αβ T cells and no change in γδ T cell number resulting in an increased γδ/αβ T cell ratio in lymphoid organs. Tg T-PPAR-β mice were partially protected against HFD-induced weight gain and exhibited decreased IR and liver steatosis independently of animal weight. Thus, in this model, a relative increase of γδ T cells was associated with beneficial metabolic effects [40]. Overall, it appears that while γδ T cells are fundamental for AT innervation and thermogenesis, the expansion of visceral WAT promotes dysregulation of its immune cell composition and likely promotes low-grade chronic inflammation [41]. The highlights of these studies described in this paragraph are presented in Table 1.

γδ T cells in human obesity

Obese adult humans have a decrease of γδ cells in the PB, which are less capable of IFNγ production and express less IL-2R and IL-7R than those of non-obese individuals but can still be activated with (E)-4- hydroxy-3-methyl-but-2-enyl pyrophosphate (HMBPP), a known potent activator of human Vγ9δ2 T cells, indicating retention of functional responses of the Vγ9δ2 subset [58]. Likewise, in a study of 10 obese women who also had diabetes, 10 obese women without diabetes and 12 controls, it was found that both PB invariant NKT (iNKT) and γδ T cells of obese patients were characterized by diminished IL-2 and IFNγ production [59]. γδ T cells are also reduced in obese patients with obstructive sleep apnea (OSA) compared to healthy controls [60]. Another study revealed that the decline may be specific for a subset of γδ T cells since obese as well as patients with colorectal carcinoma (CRC) shared a significant reduction of the Vγ9δ2/γδ T cell ratio in PB [61]. Interestingly, obese people treated with aged garlic extract (AGE) that was added to their diet, increased their γδ T cells in the PB, along with a decrease of NKT cells [62]. However, a study of 18 women with morbid obesity, revealed that only IFNγ producing NK cells but not γδ (or CD4 or CD8) T cells were positively associated with obesity, IR and hemoglobin A1C (HBA1C) levels [63]. The effect of obesity resulting in the reduction of γδ T cells in the PB may moreover, be age-dependent, since an analysis of PB of 890 children using magnetic resonance imaging (MRI) and dual-energy X-Ray absorptiometry, revealed that higher Vγ9δ2 cells were found in patients with higher fat mass index, visceral fat index, and android to gynoid ratio [64].

Similar to mice and cows, human γδ T cells may also reside in AT. For example, γδ T cells expressing Vδ1 and Vδ3 and producing IFNγ and IL-17 were increased in omental fat of patients with oesophageal cancer relative to their PB [65]. In one study, analysis of NK, NKT-like, γδ T, and Treg cell populations in VAT and blood of healthy lean subjects revealed that CD56hi NK and OX40⁺ Treg cells were more abundant in VAT with respect to blood. Conversely, CD56dim NK and total Treg cells are mostly present in the circulation, while γδ T lymphocytes were uniformly distributed in the two compartments [61]. Furthermore, analysis of AT in 120 obese and 275 controls revealed that macrophages were the source of inflammation in obese fat, whereas the relatively reduced γδ T cells that were present in AT accounted for reduced sympathetic response pathways in the AT [66]. Moreover, adiposity affects the functions of the tissue-resident γδ T cells. Thus, adipocyte-conditioned media (ACM) from obese and CRC subjects released higher levels of pro-inflammatory/immunoregulatory factors as compared to ACM from healthy lean subjects. Dendritic cells (DC), differentiated in the presence of ACM from obese (and CRC) subjects, expressed elevated levels of the inhibitory molecules PD-ligand 1 (PD-L1) and PD-L2, and showed a reduced IL-12/IL-10 ratio in response to both toll-like receptor (TLR) ligand and γδ T lymphocyte-induced maturation [67]. The highlights of these findings, pertaining to the effects of obesity on human γδ T cells, are summarized in Table 2.

γδ T cell mediated effects of obesity in diseases

It is well known that obesity negatively affects the clinical course of a number of non-malignant diseases, including psoriasis and asthma as well as that of cancer. While the primary goal of this review is to examine how obesity impacts γδ T cells in relation to their involvement in malignancy, we considered it likely that additional insights into the possible mechanisms of this three-way interaction could be gleaned by examining how γδ T cells are impacted by in the course of chronic inflammatory diseases, which are often linked to cancer [68]. Here, and in the highlighted form in Table 1, we summarize studies that have addressed the role of γδ T cells in mediating the deleterious effects of obesity in disease states in specific organs, and, finally, in cancer.

Conclusions

As summarized in graphical form in Figure 1, AT and obesity appear to have an important impact upon the function and numbers of γδ T cells in a variety of ways, that, for the most part, imply a shift toward a tumor supportive phenotype, while mitigating anti-tumoral effects. A priori γδ T cells in healthy ATs, even in the absence of obesity, appear to be skewed toward the production of IL-17, which is involved in physiologic control of thermogenesis and sympathetic innervation. Adipocytes appear to produce factors that drive DC to a more immunosuppressive state in response to γδ T cell stimulation. As obesity develops e.g., under the influence of an HFD, a more pro-inflammatory phenotype with a shift from IFNγ and toward IL-17 producing cells takes place in a variety of conditions including psoriasis, asthma, and cancer. IL-17 as discussed in the introduction is considered pro-tumoral by effects on VEGF and pro-tumorigenic neutrophils. Thus, γδ T cells in peripheral tissues of obese animals, are substantially altered: in the skin, they are reduced and lose their ability to maintain epithelial integrity. Likewise, γδ T cells are reduced among gut IEL, rendering the gut more susceptible to signals inducing chronic inflammation, thus exposing it to a pro-tumoral environment, while simultaneously driving the appearance of an enhanced γδ T17 pro-tumoral phenotype. When γδ T cells are exposed to proinflammatory conditions such as ozone in the lung, or imiquimod in the skin, obesity serves as an additional driver toward polarization of these tissue-resident γδ T cells to an IL-17 producing phenotype, predisposing these patients to harbor pro-tumoral γδ T cell populations in the skin and lung. All in all, along with the tendency of obesity, and the associated hypercholesterolemia to drive a reduction of IFNγ, these analyses suggest that a major result of obesity, is a total body reduction of the ratio of anti-tumoral IFNγ to pro-tumoral IL-17 γδ T cells as well as the induction of immunosuppressive γδ T cell phenotypes. Thus, as summarized in Figure 2, we conclude that in conjunction with numerous other immune mechanisms some of which were described above, the effects of obesity and its related conditions on γδ T cells play a relevant ancillary role in the increased incidence of cancer in obese patients. Further studies of the effects of obesity and ancillary conditions on γδ T cells will be required to completely understand how effects of obesity on these largely pro-tumoral effects could be mitigated.

Display full size

Figure 1. 

Effects of obesity on γδ T cell in specific organs. The manner in which obesity affects γδ T cells in the tissues and organs shown graphically is indicated. The curved arrows indicate feedback effects on obesity [32, 33, 35, 37–40, 45–47, 52, 58, 61–64, 66]. TH2: T helper 2

Display full size

Figure 2. 

Major mechanisms and their downstream effects on γδ T cells in organ-specific solid cancers. The mechanisms indicated on the left that are implicated in the pro-tumoral phenotypes of γδ T cells in the organ-specific cancers indicated on the right [21, 23–26, 28, 29, 46, 56, 61, 62, 64, 67]