ILC2s can indirectly regulate type 1 cytokines such as TNF-α and IL-1β, found in cases of obesity, through influencing M1 macrophage polarization. TNF-α has a significant effect on the insulin signaling pathways and directly contributes to insulin resistance and T2DM. It impairs insulin signaling through serine phosphorylation of insulin receptor substrate 1 (IRS-1) and reduction of glucose transporter type 4 (GLUT-4) expression (67). This reduction in GLUT-4 expression decreases glucose entry into cells and contributes to a lack of glucose homeostasis. The modification of IRS-1 leads to its change to an insulin receptor inhibitor in murine adipocytes, therefore abrogating the effects of insulin in adipose tissues (68). IL-1β suppresses insulin-induced glucose transport, lipogenesis, and IRS-1 phosphorylation in human and murine cell lines (69, 70). Blocking of IL-1β reduces inflammation and hyperglycemia in mouse models (71). ILC2s’ ability to polarize M2 macrophages away from an M1 phenotype would affect the production of TNF-α and IL-1β making them helpful in the treatment of T2DM. A TNF-α neutralizing antibody has also been used to ameliorate TNF-α induced insulin resistance in adipocytes (72). Cellular modulation of TNF-α and IL-1β expression through ILC2 activation and promotion of M2 macrophages may provide a longer-term treatment for T2DM.

The polarization of macrophages between M1 and M2 is the critical player in healthy adipose tissue. M2-like ATMs regulate many of the healthy functions of adipose tissue such as tissue growth and inflammation (73). While depletion of CD206+ M2-like macrophages has been shown to increase glucose metabolism in lean and obese mice (74), they function in a key niche in maintaining beige and white adipocyte progenitor populations and priming adipose tissue for healthy adipose expansion in case of high nutritional intake (73). The depletion of M2-like macrophages promoted adipogenesis responsible for observed improvements in glucose metabolism, which explained the study’s seemingly contradictory results compared to other publications showing M2-like macrophages promoting healthy metabolism (19, 47, 48). Secretion of cytokines IL-5 and IL-13, by ILC2s, regulate the generation of M2 macrophages which promote tissue homeostasis in VAT (18). IL-5 contributes to eosinophil accumulation and production of IL-4 which further aids macrophage differentiation. In cases of adipose inflammation and subsequent insulin resistance, generation of these AAMs by adoptively transferred ILC2s contributed to less adiposity and glucose tolerance in obese mice; in contrast, in vitro expanded Th2 cells were shown to not have these same effects when transferred (33). This shows that ILC2 activation is more beneficial in the regulation of adiposity and glucose homeostasis than adaptive T cells. Likewise, ILC2s are also affected by macrophages. M1 macrophages can suppress ILC2 activity through the secretion of type 2 interferon, IFN-γ, which suppresses their proliferation and secretion of cytokines (75, 76). IFN-γ deficient mice fed a HFD had smaller adipocytes, improved insulin sensitivity, and AAM shift in the VAT (77). Therefore, in this context it is very possible VAT ILC2 would have played a role in improved insulin sensitivity and M2 macrophage polarization. Controlling the balance of M1 and M2 macrophage differentiation through ILC2 cytokine secretion is crucial in the maintenance of healthy adipocytes and metabolic homeostasis.

M2 macrophages, generated by ILC2 cytokine production, can regulate a variety of homeostatic processes in the adipose tissues. M2 macrophages rely on the presence of IL-13 and IL-4 for their differentiation from macrophage progenitors (78); both cytokines are either directly or indirectly produced through ILC2s. How these M2 macrophages can induce the browning of adipocytes is hotly debated (79). Activated M2 macrophages secrete many factors which upregulate WAT beige adipocyte activity and gene expression. They secrete metabolites (80, 81), growth factors (82), and cytokines (47) which promote beige adipocytes. However, the contribution and overall effect of these factors in the grand scheme of adipose tissue homeostasis remains to be determined. Multiple papers cite the induction of beige adipocytes to be from M2 macrophage produced catecholamines (18, 19, 21, 47, 48). However, others have disputed this claim, by stating M2 macrophages do not contribute to catecholamine production (83). Future perspectives will likely determine the role of catecholamines and M2 macrophages, however, it seems that M2 macrophages affect catecholamine levels through interaction with resident nerves (84) and transport of it from peripheral locations (85). The main effector in M2 macrophage’s ability to control adipose inflammation is interleukin-10 (IL-10) and interleukin-1 receptor antagonist (IL-1Ra) (86). IL-10 seeks to downregulate immune inflammation in the adipose to achieve homeostasis through prevention of IL-6 and lipid-induced insulin resistance as observed through in vivo administration of the cytokine (87). Another major secreted factor of M2 macrophages is TGF-β. The ability of these M2 secreted factors to affect homeostatic outcomes in the adipose depends on the developmental stages of adipocytes and the immune landscape. Pre-conditioned adipose would affect the outcome of M2 macrophage-mediated effects on a case-to-case basis in the clinic.

Eosinophils, recruited by ILC2 secreted IL-5, can regulate adipose tissue inflammation. Like adoptive transfer of ILC2s into obese mice (20), adoptive transfer of eosinophils dampens systemic inflammation and improves WAT function lost with aging (88). Dysfunction of adipose tissue associated with old age mirrors that of remodeled adipose in cases of obesity (89). Systemic inflammation caused by dysfunctional adipose tissues plays a clear role in cases of T2DM. Eosinophils, the primary producer of IL-4 in WAT, have been associated with healthier glucose homeostasis (90). Their recruitment to WAT is largely regulated by IL-5 by ILC2s (18, 19). ILC2s are the major source of IL-5 compared to other lineage cells such as B, T, and NK cells, and depletion of ILC2s displays a significant reduction in adipose tissue eosinophils, not observed in the spleen or bone marrow (18). Interestingly, eosinophil-deficient mice showed reduced weight gain and body fat, however, they showed a more glucose intolerance phenotype than the wild type (91). Eosinophils play a key role in the homeostasis of adipose and proper cell function, and the inability to expand adipose during HFD underlies insulin resistance. Lipid storage then is diverted elsewhere, such as the liver, which can further contribute to glucose intolerance. Consequently, adipose eosinophils play a role in M2 macrophage differentiation, through IL-4 secretion, and subsequent production of IL-10 and TGF-β along with directly affecting adipocyte development. In the same way that eosinophils would not be active in aging adipose (89), poor trafficking of eosinophils to the adipose through dysfunctional or inactive ILC2s may play a role in the pathology of T2DM.

ILC2s have the potential to dampen inflammatory responses by mast cells through their production of cytokines. Human mast cells are suggested to be involved in the beiging process in adipose tissues (92). However, mast cells are also able to produce IL-6 and TNF-α (93) which are key in the recruitment of inflammatory macrophages and type 1 response. ILC2s have been shown to regulate mast cell activation through IL-13 secretion in dermal tissues. Although harmful in skin tissues, eosinophil infiltration following IL-5 secretion of ILC2s in the skin after the suppression of mast cells (93) may be beneficial in the adipose through the promotion of M2 macrophages by eosinophil-produced IL-4. Therefore, ILC2s may be able to simultaneously downregulate M1 macrophage inflammation mediated by mast cell type 1 cytokines and promote eosinophil infiltration and M2 macrophage polarization. Mast cells also produce prostaglandin D₂ (PGD₂) and induce chemotaxis of ILC2s through chemoattractant receptor-homologous molecule expressed on Th2 cells (CRTH2) (94, 95). Mast cells regulate ILC2 activation through the release of non-caspase proteases chymase and tryptase which cleave the activation domain of IL-33 into a more mature and potent activator of ILC2s (96). If mast cells hold an adipose tissue-specific phenotype that promotes ILC2s and adipose tissue beiging, they would be critical players in T2DM defense. Crosstalk between mast cells and ILC2s needs to be investigated fully to understand their roles in homeostasis and glucose tolerance.

There is a dynamic interaction between ILC2s and Tregs in a variety of inflammatory tissues. Tregs have long been identified for their important roles in the maintenance of VAT homeostasis and prevention of insulin resistance and glucose intolerance (97). Production of IL-33 in adipose tissue promotes ILC2 function as well as maintains Treg populations (98). Interestingly, normal expansion of Tregs was also partially dependent on ILC2s ability to express inducible T cell costimulator ligand (ICOS-L) for stimulation of ICOS+ Tregs in the adipose tissue; Tregs and ILC2s also colocalized in similar regions within the VAT. ICOS-L expression on ILC2s, and also Th2 cells, is especially high in the VAT compared to other cell types (98). In obesity-induced inflammation, sST2, a soluble isoform of IL-33R, leads to adipose ILC2 and Treg depletion ultimately leading to insulin resistance (99). It was also found that sST2 expression is the target of TNF-α adipocyte signaling. This displays the critical role TNF activation of NF-kB signaling can have on adipose homeostasis in a variety of contexts. As Tregs are master regulators of inflammation, they have also been observed to downregulate ILC2 activation in cases of allergic-type 2 inflammation in the lungs (100). However, as the inflammation is exacerbated in cases of allergy, Tregs have been shown to have a tissue-specific phenotype in the adipose (101). Due to a tissue-specific Treg phenotype in adipose, likely, ILC2s and Tregs work together to downregulate adipose tissue inflammation. Tissue-specific phenotypes of ILC2s and Tregs play a role in the maintenance of VAT homeostasis and ultimately determine glucose tolerance. In aging individuals, the number of adipose Tregs sharply increases and contributes to insulin resistance (102). The number of adipose ILC2s decrease and ILC2s become intrinsically defective due to aging; this directly contributes to thermogenic failure and insulin resistance in older individuals (103). Immunological mechanisms are likely a root cause of age-induced insulin resistance and T2DM, however, the interaction between Tregs and ILC2s in this development has not been fully investigated.

The interaction between the nervous system and ILC2s in the adipose tissue is going to be a new research area. Recent reviews have discussed the role of ILC2s in communication with the nervous system and their impact on the regulation of pulmonary disease (113). ILC2s play a similar role in adipose tissue. The sympathetic nervous system releases catecholamines in response to stimuli, such as cold temperature, leading to the direct browning of adipocytes through thermogenic gene expression. Sympathetic nerves and innervation of the WAT are required for browning of white fat through regulation of ucp1 production (114, 115). Regional variation in PRDM16 expression between adipose tissue also correlates with sympathetic neurite density (116). It has been illustrated that brown ATMs control tissue innervation and energy expenditure (117). ILC2s have been shown to co-localize with nerves expressing the neuropeptide neuromedin U (NMU) and secrete type 2 cytokines (118–120). Intriguingly, 97% of resident immune cells expressing NMU receptor 1 (NMUR1) were ILC2s, showing that this receptor is selectively expressed (120). Further studies showed that NMU increases mRNA expression of UCP1, and rats with a genetic knockout of NMU showed less calorie expenditure (121). Recently, it has been shown that NMU promotes anti-inflammatory ILC2s and Tregs to protect against arthritis (122). Other neuropeptides like vasointestinal peptide (VIP), which increases in response to caloric intake, induces IL-5 and IL-13 production by ILC2s (123). The sympathetic nervous system directly stimulates IL-33 production and adipose ILC2s in response to cold exposure. When sympathetic denervation is induced, ILC2 and eosinophil accumulation is abrogated (124). The interface between ILC2s and the nervous system in adipose tissue is a critical battleground for T2DM pathogenesis.

In cases of T2DM, the chronic release of inflammatory cytokines such as TNF-α and IL-1β contributes significantly to peripheral neuropathy (125). In the same way that this causes pain for patients, neuropathy in the adipose tissue likely contributes to a lack of thermogenesis due to sympathetic nerve damage and lack of catecholamine production. ILC2s counter the production of type 1 cytokines through the promotion of M2 macrophages (18, 19, 21), helpful in maintaining adipose homeostasis. Interestingly, previous dogmas held that catecholamine production was the main inducer of thermogenic browning of adipose via ILC2s, however, this has been challenged (83). In confirmation of both this dogma and these findings, M2 macrophages indirectly, through the protection of sympathetic nerves, and directly, due to low catecholamine production, participate in thermogenesis in WAT via ILC2s. In addition, sympathetic neuron-associated macrophages have been found to import and metabolize NE contributing to obesity and adipose dysfunction (126). These cells are recruited and activated in obesity and have increased expression of TNF-α and IL-1. In another study, certain ATMs were also found to closely associate with tyrosine hydroxylase positive nerves, and higher age mice ATMs had high expression of catecholamine degrading enzymes (127). The effect of ILC2s on these inflammatory macrophages associated with nerves in adipose tissue has yet to be investigated. In the protection of nerves, M2 macrophages have been implicated as targets for the treatment of peripheral nerve injury through their ability to accelerate tissue repair (128, 129).

The release of high levels of type 1 cytokines such as TNF-α and IFN-γ leads to neurodegeneration (130). In cases of T2DM, this may lead to suppression of thermogenesis in the adipose leading to systemic metabolic imbalance. ILC2 activation is severely diminished when sympathetic nerves are damaged (124). ILC2s are recruited to sympathetic nerves expressing NMU for the activation of immune responses (120). There, ILC2s would generate a protective type 2 and anti-inflammatory microenvironment around sympathetic nerves, which in the adipose tissue may protecting the production of catecholamines for thermogenic adipocyte differentiation. ILC2s co-localize and respond to androgenic neurons. These cells downregulate ILC2 activity through secretion of epinephrine through ILC2 expression of β₂-adrenoreceptors (131). Exploration between the crosstalk and positioning of ILC2s in the adipose at nervous system interfaces requires more investigation. ILC2s do seem to be associated in some respect with neurons, seeking to communicate and react to CNS signaling in response to changing stimuli. As other reviews indicate (132), targeting of the neuron innate immune interactions offers many new therapeutic approaches for treatments of adipose tissue dysfunction including type 2 diabetes. ILC2s can mount a defense and protect neurons from damage caused by type 1 inflammatory responses present in obesity. With restored nerve responses, the adipose tissue can increase browning associated with better metabolic outcomes ( Figure 1 ).

Adequate cell-to-cell interactions involving ILC2 activation are essential for their survival and sustained secretion of type 2 cytokines. Costimulatory molecules such as ICOS (136, 137) and GITR (20) are expressed on adipose ILC2s and healthy interaction with resident cells expressing ICOS-L and GITR-L respectively are essential for proper homeostasis. Studies have shown stimulation of GITR on ILC2s has been shown to directly protect against metabolic disturbances, such as type 2 diabetes, in obese mouse models through secretion of type 2 cytokines (20). Adipocytes were also observed to have decreased size and increased UCP1 expression associated with higher energy expenditure and restored insulin sensitivity (20). Stromal cell expression of GITR-L (144) may provide stimulation for activated GITR⁺ lymphoid cells like ILC2s, but further specific exploration is needed. DR3 stimulation of ILC2s by cells expressing the ligand TL1A leads to their expansion, survival, and enhanced effector function (28, 134, 135). Unlike GITR (20), which is only expressed when activated, DR3 is constitutively expressed in naïve ILC2s making it a promising candidate for ILC2 activation and expansion in the adipose (28, 134, 135). DR3 stimulation also provides a more significant activation of human ILC2s in comparison to GITR (28). Recently, it was found that human cells from stromal cell vascular fractions in adipose tissue induced TL1A expression in response to TNF-related apoptosis-inducing ligand (TRAIL) significantly more in comparison to other cell types (145). TRAIL is commonly associated with a variety of obesity-related diseases and plays a role in human adipocyte inflammation (146). How DR3 and TL1A interact with ILC2s in human adipose requires further investigation.

Activated ILC2s express GITR which then seeks to sustain their activation. Treatments of activated ILC2s with DTA-1, a monoclonal agonistic antibody targeting GITR, were able to prevent metabolic disturbances induced by HFD, improve glucose homeostasis, and reverse established insulin resistance in mouse models (20). Increases in lean mass percentage accompanied with decreases in VAT mass were independent of caloric expenditure through caloric intake or physical activity. Through utilizing metabolic cages, it was shown that GITR activation also increased oxygen consumption and energy expenditure. Increased expression of UCP1 and other associated calorie-burning genes were seen to be the cause of increased calorie expenditure and weight loss in the mice. DTA-1 treatment associated with higher levels of VAT M2 macrophages important for preventing metabolic disturbances (18, 21). The strongest evidence of ILC2 mediated protection against metabolic dysfunction was found through GITR^-/- mice with adoptively transferred WT ILC2s; DTA-1 treatment showed improvements in glucose tolerance solely mediated by activation of ILC2s through GITR engagement (20). Achieving specific and sustained activation of ILC2s would allow for a shift in glycolytically active tissues to a more stable and anti-inflammatory state able to prevent metabolic disturbances such as T2DM.

A more promising activator of ILC2s is DR3. Recently, a DR3 monoclonal antibody agonist, 4C12, was used in obese mouse models as a treatment to prevent metabolic disturbance and T2DM. The engagement of this receptor has shown effects in both naïve and activated ILC2s (28). Much like what was found with GITR (20), another TNF family receptor, DR3 engagement ameliorated glucose intolerance, protected the onset of insulin resistance, and reverse already established insulin resistance, but ILC2s displayed more significant effector function (28). Higher expression of adipose browning associated genes such a cell death activator CIDE-A (CIDEA), PRDM16, PGC-1α, cytochrome oxidase subunit VIIa polypeptide (Cox7a), and deiodinase 2 (Dio2) was detected in VAT lysates of DR3 agonist treated mice. The protective effect of DR3 was dependent on ILC2 secretion of IL-5 and IL-13; therefore, the observed secretion of IL-5 and IL-13 in human ILC2s with DR3 engagement in the study shows that DR3 may be a therapeutic target in human T2DM.

Adhesion molecules are some of the most important signals that regulate ILC2 function in adipose tissue. IL-33 induces the expression of ICAM-1 expression on human ILC2s and when stimulated leads to their activation. In mouse models, ICAM-1 deficiency led to attenuated ILC2 survival and development within inflamed lung tissues (147). Lymphocyte function-associated antigen 1 (LFA-1), the receptor for ICAM-1, has also been shown to be involved in the trafficking and survival of ILC2s. LFA-1-deficient mice were shown to have significantly less traffic to inflamed lung tissue; it also showed consistent expression across murine naïve and activated ILC2s unlike ICAM-1 which expression changed on activation (140). Likewise, in vivo blocking of beta-2 integrins, like LFA-1, led to significantly reduced ILC2s in blood and lung tissues (148). Determining what specific integrins regulate ILC2 accumulation in other tissues is critical. Deficiencies in adhesion molecules and their ligands consequently may be involved in the trafficking of maintenance of ILC2 populations in adipose tissue. Soluble forms of ICAM-1 have also been observed to increase in obese mouse models fed HFD leading to increased macrophage recruitment due to expanding fat mass (149). However, in other HFD fed mice, leukocyte migration to the adipose was not affected by ICAM-1 deficiency (150). Adipose tissue-resident multipotent stromal cells (MSCs), while acting as a reservoir of IL-33 for ILC2s, also stimulate them through the expression of ICAM-1 inducing proliferation and activation in LFA-1 expressing ILC2s (105). ICAM-1 depletion in the adipose MSCs leads to impaired ILC2 proliferation and effector function in the WAT; knockdown of LFA-1 on ILC2s resulted in the same findings. The ICAM-1/LFA-1 axis on ILC2s has a clear role in the regulation of adipose tissue homeostasis.

The role of immune checkpoints in the adipose tissue regarding ILC2s is becoming increasingly clear. PD-1 is directly involved in the function of ILC2s in a variety of contexts (110, 138, 139). In the adipose tissue, PD-1 is upregulated in IL-33-activated ILC2s in response to TNF-α, present in high concentrations due to obesity (110). In addition, TNF-α secretion by resident ATMs recruits and activates PD-L1^hi M1 macrophages further dampening ILC2 responses in the adipose tissue (110). In agreement with previous findings (18, 19), HFD and obesity decreased numbers of IL-13+ ILC2s, M2-like ATMs, and eosinophils (110). The number of IL-5/IL-13+ ILC2s did were unaffected by HFD in TNF-α KO mice, with the same trend being observed with M2-like ATMs (110). TNF-α injection further impaired ILC2 function and upregulation of PD-1 expression via IL-33, while depletion of inflammatory M1-like macrophages rescued ILC2s. PD-1 blockade also partially restored the ILC2/AAM/Eosinophil axis, enhanced UCP1 expression, and ameliorated glucose intolerance during a HFD (110). This study highlights that TNF signaling regulates two crucial steps in adipose tissue homeostasis via ILC2s. The first is the induction of PD-1 expression on ILC2s indirectly via IL-33, and the second is the activation of the differentiation and recruitment of M1-type macrophages from the periphery which express high PD-L1, unlike M2-like ATMs. Restoration and protection of ILC2 responses in cases of HFD and obesity are regulated by the crucial PD-1 axis.

Adiponectin, an important adipokine, was shown to restrain adipose ILC2 activation via AMPK (151). Interestingly, adiponectin has a controversial role in energy expenditure and diabetes with conflicting studies over its role as a pro- or anti-thermogenic signal. In this study, adiponectin knockout suggested an inhibitory effect on thermogenesis and a pro-adipogenic role (151). It was found adipose thermogenesis and energy expenditure were dependent on ILC2s as adiponectin levels decreased with cold temperatures (151). AMPK is activated by IL-33 in adipose ILC2s and acts as a feedback inhibition mechanism for the induction of NF-κB and therefore GATA-3 and ILC2s (151). In addition to acting on resident adipocytes and stromal cells, adiponectin was an immunological mediator stopping ILC2 mediated thermogenesis.

Cell-to-cell interactions mediated by other adhesion molecules are key in the production of IL-33 and can have an indirect effect on ILC2 functions in adipose tissues. Mesenchymal cadherin-11 directly downregulated stromal cell production of IL-33; cadherin-11-deficient mice displayed increases in IL-13 and M2 macrophage differentiation and expansion in adipose (152). It is possible that in some cases of obesity-induced inflammation there may be cadherin-11 overexpression leading to less IL-33 in adipose tissues. Lack of sustained activation of ILC2s following may lead to decreased number and effector functions. E-cadherin, the ligand for killer cell lectin-like receptor subfamily G member 1 (KLRG₁) expressed on ILC2s, has also been shown to downregulate ILC2 cytokine secretion and activation, though it was found in the context of atopic dermatitis (153). However, the expression and in vivo relevance of E-cadherin has yet to be examined in adipose tissues of patients with T2DM.

One possible contribution to insulin resistance is genetic differences in diabetic patients that lead to dysfunctional ILC2s that are unable to maintain healthy adipose homeostasis. One of the largest genetic components of this is the contribution to autophagy. It has been demonstrated that autophagy-deficient ILC2s lead to decreased cytokine secretion and increased apoptosis (154). These would both contribute to a lessened type 2 response leading to a predominantly type 1 environment in adipose tissues that would disrupt its balanced homeostasis. Intriguingly, autophagy deficiency has also been shown to disrupt innate lymphoid cell development and trafficking. Lymphocyte survival following homeostatic proliferation required autophagy (155). ILC2 precursors, deficient in autophagy and being exposed to a highly inflammatory cytokine milieu, may be unable to reach their target tissues due to induction of apoptosis for failing to meet their energetic needs. Obese patients have been shown to have altered autophagy which may contribute in some patients to diabetes. In these cases, it seems that autophagy can be either enhanced or attenuated in different patients contributing to their obesity; therefore, the heterogeneity of the human population may contribute to some obese patients developing T2DM or not (156). Other polymorphisms involved in ILC2 function may also affect their ability to fight T2DM. As observed with ST2 deficiency in mouse models (157), brown adipocytes displaying polymorphisms in their ST2 gene may be a cause of inadequate ucp1 protein product. This could have been mediated by the ST2 deficiency in ILC2s leading to their inability to induce UCP1 expression like observed in Galle-Trager et al. (20). Although most studies of ILC2s focus on the effects of cell autophagy on lung inflammation (158), there are still questions as to how autophagy in adipose tissue ILC2s can contribute to T2DM pathogenesis. Likely, genetic differences in the human population, especially in all-encompassing genes such as autophagy-related genes, show connections to obesity and ultimately may contribute to the pathogenesis of T2DM.

Interferon signals activate STAT1 for the transcription of genes for a more specific response to viral infection. A patient is nearly four times more likely to develop T2DM if they are infected with hepatitis C virus (HCV) (159). Successful treatments of HCV in T2DM patients displayed improvements in glycemic control and resulted in less insulin use by patients (160). Innate immune cells such as dendritic cells are responsible for the production of IL-12 in response to HCV infection through TLR-3 signaling leading to NK cell production of IFN-α (161). High IL-12 results in ILC1-like ILC2s with the expression of T-bet (162). These ILC1-like ILC2s produce IFN-γ which further diminishes committed-ILC2 effector function (75, 76, 98). Meta-analysis has shown that HCV infection is associated with an increased risk of T2DM independent from associate liver disease (163). The true cause of this association has yet to be observed, however immunological mediators of the pathogenesis of HCV are likely involved such as ILC2s. In addition, any production of IFN, type 1 or IFN-γ, due to HIV infection may be a causative agent in T2DM comorbidity; patients infected with HIV are also up to four times more likely to have T2DM (164). Plasmacytoid dendritic cells (pDCs) have been shown to produce copious amounts of IFN-α in response to HIV (165). The link between viral infection and the diagnosis of T2DM is still being investigated.

Interferon originating from peripheral locations can lead to restricted Th2 and Th17 responses, including that from ILC2s (166). STAT1 deficiency leads to a lack of proper response to pathogens, but overactivation of the transcription factor can lead to autoimmunity and sustained inflammation due to enhancement and positive feedback in IFN-α/β signaling (167). Genetic gain of function in STAT1 has not been investigated in metabolic diseases such as T2DM, although interferon would lead to decreases in ILC2s (75, 76, 98). Classically, the survival of ILC2s relies on stimulation by IL-2 and IL-7 as well as their functional receptors. Genetic defects in these signaling pathways leading to STAT activation may contribute to ILC2s’ inability to maintain adipose homeostasis and insulin sensitivity. STAT6 activation through IL-4 receptors (IL-4R) plays a key role in homeostasis and cytokine production in ILC2s. Mice deficient in STAT6 have been shown to have impaired numbers of IL-13 producing ILC2s (168), a phenotype usually displayed in IL-4R deficient mice. More genetic studies in humans are needed to quantify the effects of polymorphisms of specific transcription factors for their implications into ILC2 mediated adipose inflammation and T2DM.

Though adipose ILC2s are considered tissue-resident cells with low-level replenishment from the periphery (12), in cases of inflammation the trafficking of ILC2s changes (13). The ability of ILC2s to traffic to inflamed tissues, such as adipose tissues, is essential for the prevention of insulin resistance. Overall, more studies need to be done to fully understand the process of ILC2 trafficking to multiple peripheral tissues. Expression of CRTH2 and its ligand PGD₂ regulates chemotaxis of ILC2s to the lungs in the context of allergic asthma (94, 95, 209). Stimulation of mineralocorticoid receptors, often associated with blood pressure dysregulation, has been shown to increase adipocyte PGD₂ synthase (210), which could regulate ILC2 accumulation. However, the chemotaxis of ILC2s to diabetic active tissues still requires more research. Other receptors such as cysteinyl leukotriene receptor 1 (CysLT1R) on ILC2s and leukotriene D₄ (LTD₄) concentrations may also regulate chemotaxis to inflamed tissues (211). Leukotrienes (LT) like LTC₄, LTD₄, and LTE₄ all were shown to induce migration, cytokine secretion, and reduce apoptosis in ILC2s; LTE₄ was specifically shown to enhance effects of PGD₂, IL-25, and IL-33 (212). How these chemotactic markers work in trafficking ILC2s to adipose is still up for debate.

One recent subject of interest has become mature ILC2 trafficking to and between diverse tissue sites. ILCs are largely found as tissue-resident cells (11), but through migrations, they can regulate systemic homeostasis. IL-25 has been shown to regulate the trafficking of KLRG1^High inflammatory ILC2s (iILC2) from the small intestine to the lungs through differential expression of sphingosine 1-phosphate (S1P) receptors (13). This traffic was found to be both to lymphatic and non-lymphatic organs, with a substantial number of ILC2s originating from the gut. Interestingly, the results indicated that intestinally derived iILC2s infiltrate the bone marrow where they can have effects on ILC2 progenitors (13). Although IL-25 has been shown to stimulate a type 2 inflammatory response associated with M2 macrophage polarization adipose tissue of mice (213), natural sources of IL-25 in the adipose remain elusive. However, injections of IL-25 in obese mouse models display enhance UCP1 expression, M2 macrophage polarization, and overall improvements in glucose homeostasis (84), showing that activation of ILC2s in adipose may provide a route of treatment for T2DM. Disrupted trafficking of ILC2s between organs in cases of T2DM inflammation still needs to be further investigated.

Depending on the type of immune cell their destination and outcome due to the cytokine and chemokine milieu in the blood can be completely different. ILC2s exiting the bone marrow are met with a variety of cellular signals which dictate their outcomes ( Figure 3 ). In cases of systemic inflammation, such as obesity-induced diabetes, pro-inflammatory cytokines can be present in the blood and concentrated in fat tissues as activated adipose tissue macrophages (ATMs) produce tumor necrosis factor-alpha (TNF-α) and interleukin-1β (IL-1β) (111). These cytokines lead to the activation of ILC2s for the production of cytokines IL-5 and IL-13 (162, 214), however, due to the plasticity of ILC2 populations, the strong presence of these cytokines can also lead to the conversion of these cells into ILC1 phenotypes (162) (ex-ILC2) to sustain the type 1 immune response. In cases of obesity, ILC2 progenitors exiting the bone marrow may be met with strong signals from type 1 cytokines leading to their differentiation into an ILC1-like phenotype. Interestingly, IL-12, a requirement for T-bet⁺ ILC2s (162), is elevated in obese patients in a variety of studies (215–217) showing that obese patients may show a suppression of the ILC2 phenotype. ILC1s are directly linked to the pathogenesis of T2DM through inducing adipose fibrosis and M1 macrophage activation (170). Adipose-resident ILC1s also promote insulin resistance in cases of obesity through the production of IFN-γ necessary for pro-inflammatory macrophage polarization (169). Production of TNF-α by ILC1s would lead to further exacerbation of insulin resistance which has a variety of downstream consequences for insulin signaling. These directly affect serine kinases (IKK, JNK, S6 Kinase) and mTOR which are involved in attenuating insulin signaling and healthy response to the hormone (86).

Interestingly, interferons like IFN-γ and interleukin 27 (IL-27), have been shown to directly restrict ILC2 activation, cytokine production, and increase apoptosis (75, 76, 98). Activated natural killer cells (NK cells) are often the main source of IFN-γ in the adipose tissue and its secretion leads to pro-inflammatory macrophage accumulation. NK cells are directly activated in cases of obesity through the upregulation of NK cell-activating receptors on adipocytes; more specifically through the expression of NKp46, known as natural cytotoxicity triggering receptor 1 (NCR1) in mice (218). They also directly contribute to insulin resistance in obesity (219). Circulating cells like pDCs produce a copious amount of IFN-α which has been shown to regulate ILC2 functions. IFN-α from pDCs has been shown to directly suppress cytokine production, proliferation, and increase apoptosis in ILC2s (220). Circulating cells such as pDCs therefore may play a role in ILC2 progenitor development through IFN-α signaling. Signal transducer and activator of transcription 1 (STAT1), a downstream transcription factor of IFN-γ signaling, plays a key role in the control development of adipocytes and the ability of interferon to induce insulin resistance (206). Therefore, interferon production is a key player in adipose inflammation and homeostasis, not only circulating ILC2 progenitors. High circulating interferon would polarize ILC progenitor towards an ILC1 phenotype. Halting other phenotypes of innate lymphoid progenitors and promoting ILC2s is essential in maintaining healthy adipose tissue homeostasis and preventing metabolic remodeling leading to insulin resistance.

Trafficked naïve ILC2s to the inflamed adipose are key in the regulation of the cytokine profile and outcome of the tissue. Maintenance of arrived ILC2s is dependent on pro-survival signals. One of the most potent activators of not only activated but naïve ILC2s is DR3 (28, 134, 135). In contrast to GITR (20), DR3 activates ILC2s through both canonical and non-canonical NF-κB pathways (28). It is therefore plausible that the observed increase in mouse VAT ILC2 activation, when stimulated with DR3 agonist, is due to this doubled activation of the NF-κB pathways. Treatments seeking to maintain ILC2s in the adipose through activation of DR3 may be a possible therapeutic route for T2DM treatment. Antigen-presenting cells in the VAT expressing TL1A (221, 222) also may be able to specifically stimulate newly trafficked ILC2s in the adipose tissues and therefore would be able to regulate their numbers. Although TL1A expression has been explored in a variety of different contexts (221), the role of TL1A in adipose has yet to be fully explored. In one study, TL1A-deficient HFD mice showed decreased adiposity, glucose tolerance, but also reduced expression of UCP1 (223). It was also found that TL1A deficiency also reduced the abundance of ILC1s. Interestingly, these findings indicate that the effect of TL1A is dependent on the context. Obesity conditioned adipose may be receptive to DR3 agonists. However, in a predominant type 1 inflammatory environment, the presence of a DR3 agonist may be detrimental as it would stimulate more ILC1s than ILC2s. Therefore, the timing of these treatments and understanding the inflammatory phenotype of the patient is extremely important. The ability to shift the adipose microenvironment towards a more Th2 cytokine profile and predominant ILC2s would likely prove helpful if treatments utilizing DR3 agonists were to be considered. Early-stage treatment for T2DM would therefore require DR3 antagonization for increases in glucose tolerance and insulin sensitivity, followed by DR3 activation for the sustained maintenance of adipose trafficked ILC2s and increase in UCP1 expression and adipose browning. Treatments seeking to traffic and maintain ILC2s in the adipose tissue will play a key role in future treatments of T2DM and conditions of adipose deregulation.

Other genetic factors leading to the release of interferons can regulate populations of ILC2s through dampening their effector function. A recent hypothesis in the pathogenesis of T2DM is that a leaky gut leads to systemic inflammation through the release of interferons. In addition to the leaking of short-chain fatty acids and bile acids leading to insulin resistance, the bacteria and LPS leaking out of the gut leads to the release of a variety of inflammatory cytokines from immune cells (224, 225). Cytokines like IFN-α/β can leak into the bloodstream and spread peripherally negatively affect ILC2 number and function (75, 76, 98). The microbiota is being explored in a variety of contexts due to its role in the initiation of gut inflammation. ILC2s in the gut have many interactions with bacteria, however, their contribution to T2DM has yet to be explored fully. Commensal microbiota has been observed to impact epigenetic regulation and gene expression of ILCs (226). Consequently, there are likely genetic and external components of the gut microbiota that may affect the inflammatory outcomes of ILC2s and T2DM on a more systemic scale. There is a new and upcoming role of microbiota in the understanding of immunity and the development of seemingly unrelated diseases like T2DM.

Relatively new in the study of innate biology, ILC2₁₀s, IL-10 secreting ILC2s, have been implicated in a variety of inflammatory contexts (227). ILC2s have been observed to secrete IL-10 in lungs (228–230) and gut tissues (231). It was recently discovered that ILC2₁₀s express transcription factors cMaf and Blimp-1 for IL-10 secretion in the context of allergic lung inflammation (229). IL-10 production by ILC2s has also been inducible by NMU (231), emphasizing the importance of the sympathetic nervous system in ILC2 regulation. Although it is currently unknown, these regulatory ILC may play a role in the downregulation of inflammation in the adipose. As the relationship between type 1 and type 2 inflammation is dichotomous, secretion of IL-10 and Treg cells ultimately leads reduced overall immune activation and complete homeostasis. It is reasonable that ILC2₁₀s may play a key role in more long-term outcomes of T2DM patients and downregulation of all types of inflammation in adipose tissue. Interestingly, they have also been shown to reduce in vivo eosinophil recruitment (228). Discovery of this very plausible population of ILC2₁₀s in cases of extreme adipose inflammation is needed as the expansion of this population would likely provide a treatment option for T2DM.

The timing and development of obesity in patients can change the outcome of T2DM and can affect its pathogenesis. For the healthy expansion of adipose tissue, there is a need for expansion of not only adipocytes but also immune cells (232). Healthy adipose expansion requires proper regulation of the endothelium in terms of adhesion molecules that enhance immune trafficking to keep up with the growing tissue as well as vascular growth factors for oxygen distribution. In unhealthy adipose tissue expansion, adipocyte hypertrophy and cellular stress lead to inflammation in the form of M1 and NK cell accumulation (56). In addition, angiogenesis decreases while fibrosis and hypoxia increase. The healthy adipose expansion increases M2 macrophages and Tregs (56). It is likely that in the case of rapid adipose expansion, endothelium adhesion molecules are unable to traffic enough ILC2s to keep up with a quickly expanding inflammatory environment. In addition, stromal cell IL-33 is unable to stimulate tissue-resident ILC2s. Sources of ILC2 activation have to expand, including adhesion molecules and stromal cell populations, with the growing adipose. Collagenous receptors such as LAIR-1 on ILC2s may also be involved in adipose expansion (233). There is likely is a lag time between when adipose is under rapid expansion and when ILC2s react. In case of extreme expansion, this lag time would lead to an unrecoverable ILC2 population. The rate at which the adipose expands would therefore have a direct effect on the immunology of the adipose and affect outcomes of T2DM in patients.