Macrophages comprise the most well studied immune cell population in the AT. While the AT literature mostly classifies macrophages into the simple M1/M2 (pro/anti-inflammatory) subsets, macrophages are highly plastic, and many more subsets may be described according to a wide series of markers and functions (60, 61).

Macrophages within WAT and BAT present different profiles (62). BAT macrophages are generally profiled as the M2 subset (63) and have been proposed to support thermogenic functions (27). In addition to their role in BAT, M2 macrophages were also reported to facilitate browning by clearing out dead adipocytes and favoring the recruitment of PDGFRα⁺ progenitor cells after β3-adrenergic receptor activation (64). A later study showed that mice receiving adipose stem cell-derived exosomes induced M2 macrophage activation, favored WAT browning and improved metabolism (65). In another study, the suppression of M2 polarization affected BAT activity and WAT browning in obese mice (35). Conversely, other studies linked the M1 phenotype induction with loss of browning in WAT and impaired BAT thermogenic function (36, 37). Thus, M2 polarization, browning, and BAT functions are positively correlated (38, 39). However, a recent study claimed a subset of M2 macrophages to partially hijack beige progenitor proliferation and browning (66). Overall, the precise link between macrophage alternative activation and thermogenesis still remains to be formally established, and it is important to highlight that a deep categorization of all macrophage populations was missing in these studies, raising the possibility that further distinct subpopulations of macrophages mediate different effects, potentially explaining some discrepancies.

M2 macrophages were initially reported to sustain adaptive thermogenesis through catecholamine production in response to cold (67, 68), bringing a new perspective in the study of TAT modulation through immune cells. However, recent studies disclaimed this fact, as no changes in energy expenditure were found between wild-type, Ucp1^-/- and Il4ra^-/- mice, nor were AT-resident macrophages (ATMs) expressing the catecholamine synthetizing enzyme, tyrosine hydroxylase (69). In fact, a subpopulation of macrophages, the so-called sympathetic neuron-associated macrophages (SAMs), has been suggested to metabolize catecholamines via the expression of the norepinephrine (NE) transporter solute carrier family 6 member 2 (SLC6A2) and the degradation enzyme monoamine oxidase A (MAOA) (40). Macrophages were also found to degrade NE via NLRP3 inflammasome activation (70). These conflicting results could rise from the fact that macrophages sustain sympathetic innervation (71, 72), and although not directly synthetizing it, they might contribute to the local titers of NE produced by sympathetic neurons present in TAT by directly communicating with them (73, 74). More recently, a study pointed out CX3CR1⁺ SAMs as a potential source of IL-27, which has been shown to contribute to thermogenesis and energy expenditure (75). Besides, a novel study proposed a subpopulation termed cholinergic adipose macrophages (ChAMs) to secrete acetylcholine and regulate thermogenic activation of beige fat (41), adding another layer of complexity to the global contribution of macrophages to sympathetic activation. The sufficiency of M2 macrophages in promoting browning independently of sympathetic neuron involvement was described in a Fasn ^-/- mouse model (76). Furthermore, macrophages can also express uncoupling protein 1 (UCP1) and aid on beige AT remodeling after cold exposure (77).

Lineage tracing human studies suggested that WAT macrophages could dedifferentiate towards pre-adipocytes and vice versa (78). If applicable to the thermogenic depots, this could become a valuable approach to increase the portion of energy-burning cells within TAT. In other tissues such as the heart, macrophages have been proposed to be essential for mitochondria turnover leading to severe alterations in cardiac function and metabolism when ablated (79). This opens up a new horizon for alternative roles of macrophages in TAT and could lead to new therapeutic approaches.

Eosinophils and type 2 innate lymphoid cells (ILC2s) are usually present in the mucosas and act during helminthic infections and allergies (80, 81), although their functions are more diverse and vary depending on their location in the body (82). In adipose depots, eosinophils express high levels of Siglec-F and produce IL-4 and IL-13 to induce M2 macrophage polarization (42). Genetic ablation of eosinophils or the IL4/13-IL4Rα-STAT6 pathway has been linked to impaired cold-induced browning as a result of loss of macrophage M2 polarization (67). Therefore, the overall contribution of M2 macrophages to TAT functions and browning may be regulated upstream by the presence of eosinophils and/or IL-4/13 signaling. Type 2 immune signaling could hence be a promising approach for the treatment of metabolic diseases. This strategy was tackled in a study which tried to counteract the HFD-induced eosinopenia by treating mice with helminth antigens (83). Despite increased eosinophils in WAT, the treatment neither did affect eosinophil numbers in BAT nor induced WAT activation or browning. Considering that immune cell profiles vary depending on their environment, the immune response against helminths might have not been enough to induce the same eosinophil profile required for browning in vivo. However, advancements in understanding eosinophil plasticity could aid on harnessing type 2 immune responses against metabolic disorders.

Eosinophils infiltrate BAT in response to CCL-11 and IL-5 (43, 83, 84). ILC2s are the major source of IL-5 and their functions are negatively affected during obesity (43). ILC2s sustain adipose eosinophils and macrophages through IL-5 and IL-13 production (43) albeit it has been proposed that ILC2s alone could be sufficient to elicit browning and promote metabolic homeostasis independently from eosinophils (50). Upon activation by IL-33, ILC2s promote PDGFRα⁺ adipocyte precursor proliferation and commitment to the beige lineage via IL-4Rα, which overall culminates in increased beige fat mass and improved energy expenditure (51). A new study also revealed an intercoordinated mechanism between ILC2s and eosinophils to enhance sympathetic innervation in WAT (85). Detailed mechanisms by which ILC2s contribute to beige adipogenesis, and metabolism have been described elsewhere (86, 87). Overall, type 2 immune cells are able to directly and indirectly contribute to the positive functions of TAT in energy homeostasis. While the underlying mechanisms are emerging, further research is needed to decipher the complex array of communications between type 2 immune cells and adipocytes.

Regulatory T cells (Tregs) constitute a CD4⁺ T cell subset that highly expresses the forkhead box P3 (FOXP3) transcription factor and are crucial for the maintenance of self-tolerance and immune homeostasis. AT resident Tregs were only discovered by the end of the last decade (88), however, large-scale research allowed adipose Tregs to be one of the best characterized along all tissues. AT resident Tregs express PPAR-γ (89) as well as a distinct T cell-receptor (TCR) repertoire compared to other Tregs in the lymph node (88). Correspondingly to many other immune cells, Treg characterization and relevance have been more largely described in WAT (90), albeit it is predictable that these cells also play major roles in TAT.

BAT Tregs harbor a specific transcriptomic signature different from splenic Tregs or T conventional (T_conv) cells in all tissues, including T_conv cells in BAT (46). Interestingly, this signature is also different from WAT Tregs, which incites to question whether these cells perform different functions in WAT and TAT. Tregs have been proposed to maintain BAT identity and promote browning (46, 91), and a recent study showed that inflammatory signals mediated by the insulin receptor in visceral WAT and BAT Tregs give rise to detrimental outcomes in the regulation of diet- and age-induced metabolic syndrome (92). Other studies in mouse and humans also showed the presence of resident naïve CD4⁺ T cells that skew towards a regulatory phenotype under certain stimuli (i.e cold), describing them as cold-inducible Tregs (44, 45). Cold-inducible Tregs mainly increase in BAT and beige AT and their depletion automatically affects the thermogenic identity of the tissue. Inversely, UCP1 ablation impairs Treg induction (44). Therefore, each cell type relies on one another to sustain adaptive thermogenesis. Still, whether Treg-adipocyte communication happens directly upon cold exposure or trough complex signaling events that lead to secretion of molecules needs to be investigated.

On the other hand, Tregs resemble the boring parents that shut down the party when it is time to go home. But how are these cells prevented to restrict too harshly? The AT harbors such a wide array of immune cells that it even allows to regulate the regulatory cells. This is the case for invariant natural killer T (iNKT) cells, a singular adipose resident immune cell subset, which tightly modulates the proliferation and suppressor function of adipose Tregs through the secretion of IL-2 (54). Besides, α-galactosylceramide-activated iNKT cells upregulate fibroblast growth factor 21 (FGF-21), which subsequently promotes adipocyte browning (55). Another subset of T cells, the so-called unconventional γδ T cells, has also been shown to support Treg cell homeostasis and tissue innervation through IL-17 signaling (52, 53).

Elucidating all the immune populations in TAT and their contribution to metabolism has become an attractive field of research. More immune populations have been recently described in TAT. Monocytes have been shown to support BAT homeostasis by promoting tissue expansion (56). B cells comprise 20 to 30% of leukocytes in BAT, and their number is further increased upon diet-induced obesity (48). Along with T cells, B cells have been suggested to negatively influence thermogenesis via adipocyte IL-10 signaling (49). CD8⁺ T cells have also been suggested to negatively influence beige adipogenesis through IFN-γ secretion in a lymphocyte deficient mouse model (47). Mast cells are also present in TAT and communicate with adipose cells through factors like histamine or serotonin, albeit whether they promote or hamper WAT browning remains to be clarified (57–59). Accordingly, further studies on these populations will be required to sustain the presented findings. More recent studies using single-cell or nuclei RNA sequencing approaches have shown still undescribed immune cell populations, such as neutrophils or dendritic cells, to be present in TAT of mice (56, 76, 93) and humans (93). It is therefore encouraging to expect that the roles of more immune cell populations in thermogenic depots will be unveiled in the coming future. Furthermore, it will be important to not only consider the ability of TAT immune cells to influence tissue identity and function, but also to modulate energy expenditure, following appropriate guidelines (94, 95).