Sympathetic Innervation of White Adipose Tissue: to Beige or Not to Beige?

Abstract

Obesity research progresses in understanding neuronal circuits and adipocyte biology to regulate metabolism. However, the interface of neuro-adipocyte interaction is less studied. We summarize the current knowledge of adipose tissue innervation and interaction with adipocytes and emphasize adipocyte transitions from white to brown adipocytes and vice versa. We further highlight emerging concepts for the differential neuronal regulation of brown/beige versus white adipocyte and the interdependence of both for metabolic regulation.

Introduction

Adipose tissue biology made huge progress in the last decades, largely driven by the raging obesity epidemic and the need for a better understanding of how and why adipose tissue is able to expand tremendously and why chronically increased fat mass is associated with metabolic diseases like type-2-diabetes (T2D), cardiovascular problems, and most recently increased infection vulnerability for Covid-19 symptoms. The historic view of adipose tissue as a mere energy storage site has been replaced with the understanding of adipose tissue as an important endocrine organ, with leptin and adiponectin as the most prominent hormones, as well as many other relevant adipokines (1, 2). The obesity research field is now keenly aware that adipose tissue plays a central role in metabolic disease, which is promoted by excess fat accumulation in overweight and obese individuals, but also by the lack of adipose tissue in individuals with lipodystrophy (3, 4).

It is well accepted that adipose tissue is under neuronal control from the sympathetic nervous system (SNS) and sensory nerves provide feedback from the adipose tissue to the central nervous system (CNS) (5), and despite initial controversy (6), it is generally accepted that adipose tissue does not receive any parasympathetic innervation. Functionally, the sympathetic tone in adipose tissue promotes lipolysis and induces thermogenesis in brown adipocytes via activation of uncoupling protein 1 (UCP1). Another prominent sympathetic function is the induction of thermogenic features in white adipose tissue (WAT). This switch of white to brown adipocytes (termed “beige” or “brite” adipocytes) is known as “beiging” or “browning” (7–9). The interface of sympathetic tone and adipocyte function is mediated by adrenergic receptors (AR) where the sympathetic transmitter norepinephrine binds and mediates lipolysis, thermogenesis, and WAT beiging.

There is little doubt that WAT beiging is present in humans and rodents, even though the functionality and capacity of beige fat for body weight regulation in humans are debated (10). This review will focus on emerging concepts that address specific aspects of the sympathetic innervation and regulation of adipocyte function. We will discuss existing discrepancies in the literature and highlight gaps in the understanding of the neuro-adipocyte interface with regards to adipose tissue beiging and thermogenic coupling and the importance within the larger context of physiological adaptations in health and disease.

CNS Sympathetic Circuits and Adipose Tissue Innervation

The anatomy of adipose tissue innervation circuits involves hypothalamic and brainstem neurons, preganglionic spinal cord neurons, and postganglionic sympathetic chain ganglia neurons (FIGURE 1). These circuits have been mapped from their sympathetic endings in interscapular brown adipose tissue (iBAT) and inguinal WAT (iWAT) to higher order neurons in the brain with the use of retrograde, transsynaptic viral tracers (11–14). Several hypothalamic sites (preoptic area (POA), dorsomedial hypothalamus (DMH), paraventricular hypothalamus (PVH), and pro-opiomelancortin (POMC) neurons were identified with multisynaptic connection to iBAT or iWAT and play an important role in different aspects of thermoregulation (body temperature, fever, metabolism, and body weight) (13, 15–22). Some sites and neurons, like the ventromedial hypothalamus (VMH) or agouti-related-peptide (AgRP) neurons, may be too far upstream in the circuit to be detected by viral tracing methods, but their involvement in energy expenditure regulation is well established (23–27). The VMH has received particular attention for its role in sex-specific metabolic control (24). However, we currently lack a clear understanding if and how these hypothalamic sites interplay during a given physiological challenge like cold temperature, fasting, overfeeding/obesity, or fever (as recently reviewed Ref. 22) and if they also modulate iWAT beiging. Although select neuronal populations are known for their robust impact on iWAT beiging (28–31), if and how these neurons connect with thermoregulatory circuits remains unclear.

FIGURE 1.

Adipose tissue innervation scheme This scheme highlights the emerging concept of differential sympathetic tone and innervation density in select adipose tissue depots [e.g., interscapular brown adipose tissue (iBAT) and inguinal white adipose tissue (iWAT) and beige islands]. Brain circuits connect neurons either directly (e.g., POMC, PVN, and RPa neurons) or indirectly (e.g., POA, DMH/DHA, POMC, and AgRP neurons) with presympathetic neurons in the spinal cord (1 and 2). The differential location of iBAT vs. iWAT innervating presympathetic neurons is a promising interface for differential regulation of postganglionic sympathetic tone to select adipose tissue depots (3 and 4). POA, preoptic area; PVH, paraventricular hypothalamus; DMH/DHA, dorsomedial hypothalamus/dorsal hypothalamic area; POMC, pro-opiomelanocortin; AgRP, agouti-related peptide; RPa, raphe pallidus; StG, stellate ganglion; T, thoracic level; IML, intermediolateral bundle; NE, norepinephrine.

Furthermore, in contrast to detailed studies on hypothalamic circuits, very limited work has addressed how hypothalamic neurons communicate with adipose tissue innervating preganglionic neurons in the spinal cord and if they may differentially regulate iBAT and iWAT. We recently performed a detailed mapping study to identify the full extent of the preganglionic inputs to iBAT and iWAT (32, 33) and highlighted the stark anatomical dissociation of pre- and postganglionic inputs to iBAT versus iWAT (FIGURE 1). Hypothalamic neurons can connect directly or indirectly (via premotor neurons) and innervate preganglionic, cholinergic neurons in the intermediolateral column (IML) of the spinal cord. A premotor site for sympathetic iBAT innervation is the raphe pallidus (RPa) in the mediobasal brainstem (FIGURE 1), and hypothalamic neurons in the PVN, POA, and DMH regulate iBAT thermogenesis via the Rpa (16, 18, 34). Anatomical evidence also indicates the RPa as premotor input to iWAT sympathetic circuits (21). Hypothalamic neurons that project directly to the spinal cord are found in the parvocellular PVN (e.g., oxytocine and vasopressin expressing neurons) (35), select POMC-expressing neurons (11, 36), and orexin neurons in the lateral hypothalamus (37), even though their importance for adipose tissue regulation is unclear. However, projections of these hypothalamic neurons are consistent with the spinal cord levels where iWAT and/or iBAT innervating preganglionic cell bodies are located (11, 36).

The melanocortin system is an important component of sympathetic adipose tissue innervation and includes POMC- and AgRP-expressing neurons with opposing effects at the melaconortin-4-receptors (MC4R). MC4R-expressing neurons are found in the spinal cord-projecting hypothalamic neurons that connect to iWAT (13) and iBAT (11) sympathetic circuits. Bartness and colleagues (38) first introduced the concept of differential sympathetic activation of iBAT versus iWAT and demonstrated that physiological adaptation to energy deficiency states (fasting, glucoprivation) increased norepinephrine turnover (NETO) only in iWAT but not iBAT. Similarly, energy deficiency causes differential activation of AgRP neurons, while inhibiting POMC neurons (22), which causes an overall inhibition of MC4R signaling and decreased thermogenesis. Conversely, MC4R stimulation increased NETO in both iWAT and iBAT (39) and is well known to increase thermogenesis.

MC4R is expressed in hypothalamic and IML neurons (13, 14, 40–42), but the relative impact of these neurons for adipose tissue innervation and beiging has not been explored. However, IML-specific MC4R signaling significantly impacts body weight, energy expenditure, and glucose homeostasis and is uniquely innervated by POMC neurons and lacks AgRP neuronal inputs. Melanocortin neurons are also highly responsive to the adipose tissue-derived hormone leptin, and circulating leptin levels are controlled by the sympathetic tone to adipose tissue (43–47). Furthermore, leptin levels impact other hormones like adiponectin, glucocorticoids, and insulin (2, 48–50). Recent work further suggests that leptin action on POMC and AgRP neurons differentially affects iBAT versus iWAT sympathetic tone (51). These data suggest a disconnect of POMC leptin action from sympathetic iWAT, but not iBAT, tone and contribute to an emerging concept that the relative sympathetic tone in iBAT versus iWAT might be an important determinant of adipose tissue thermogenic function.

Adipose Depot Sympathetic Innervation, Proliferation, and Beiging

Adipose tissue depots have enormous flexibility with adipocytes that can expand in size (hypertrophy), increase in cell numbers (hyperplasia/proliferation), transition their metabolic features from white to brown/beige adipocytes (beiging/trans-differentiation), and generate de novo beige adipocytes (52–55). These phenomena are of great interest, as they are largely associated with increasing (hyperplasia/proliferation) or decreasing (beiging/trans-differentiation) body weight and as such are important mechanisms to understand illnesses associated with body weight changes such as obesity or cancer- and other illness-associated weight loss (for detailed reviews see Refs. 55–58). However, it is important to point out that changes in adipose tissue size and metabolic efficiency are part of normal physiology and adaptations to daily or seasonal changes (e.g., ambient temperature, food availability, stress, sleep) and robust seasonal changes in body adiposity and insulin sensitivity are not inherently associated with illness (59–62).

Differential Innervation within Adipose Tissue Depots

An emerging concept is that overall innervation density is tightly associated with adipocyte status (FIGURE 2), and high innervation density goes together with high brown/beige adipocyte density. This is true when comparing iBAT to adipose tissue depots that are most prone (iWAT) or least prone (epididymal WAT) to beiging (63, 64) or when comparing select regions within iWAT (dorsolumbar versus inguinal portion) (63, 65), with the differential appearance of beige islands within the iWAT (33, 65).

FIGURE 2.

Complex communication at the adipocyte/innervation interface Schematic diagram showing the neuronal axis from central nervous system (1) via sympathetic nerve (2) to the differential innervation density (3) of adipocytes. Sympathetic activity and adrenergic signal transduction (4 and 5) are actively involved to modulate innervation density and adipocyte status (6) from brown adipocytes (either in interscapular brown adipose tissue or beige islands of interscapular white adipose tissue) to white adipocytes. Vice versa innervation density is dependent on adipocyte status. Thus the highlighted complex communication interface for neuron/adipocyte shows an emerging concept that modulations at any level in this axis will impact innervation density and adipocyte status accordingly [7: humoral response; 8: cross talk].

Recent technological advances with adipose tissue clearing have given us unprecedented insight into the sympathetic innervation network of iWAT and the rich innervation of iWAT has been emphasized (63). However, only axonal varicosities, but not synaptic branches, have norepinephrine-filled vesicles that are released upon neuronal activation and are the functional unit of the sympathetic nerve. These axonal varicosities densely innervate adipose tissue arteries and are found in close proximity to adipocytes and even make close synaptic contacts (20 nm, no basal lamina), at least among brown adipocytes (66). Axonal varicosities are associated with increased branching toward beige adipocytes and dense varicosities are found close to UCP1-expressing beige adipocytes, while adjacent white adipocytes are sharply distinguished by sparse varicose endings within the same adipose depot (33) (FIGURE 2). Thus the interaction of adipocyte and sympathetic varicosities seems highly relevant to maintain adipocytes in their beige/brown state, while denervation initiates hyperplasia/proliferation and whitening of beige and brown adipocytes (67, 68).

A helpful working model is shown in FIGURE 2, depicting the brain and preganglionic circuits as modulator of sympathetic activation (bullets 1 and 2). The innervation density (bullet 3) and thus sympathetic norepinephrine release and signal transduction via adrenergic receptors (α- and β-AR, bullets 4 and 5, respectively) is an important component to recruit beige adipocytes, to maintain brown/beige adipocyte status (bullet 6), and to prevent adipocyte hyperplasia/proliferation. This is supported by surgical or chemical denervation of either iBAT or iWAT, resulting in reduced expression of the brown adipocyte marker UCP1(68) as an indicator of de-differentiation, while promoting adipocyte hyperplasia/proliferation (13, 69, 70). Although adipose tissue denervation does not cause obesity, a disruption of adipose tissue innervation dynamics may likely contribute to obesity development with predominant adipocyte hypertrophy and proliferation, and it is critical to understand the mechanisms that regulate the dynamic changes in adipose tissue innervation, activation and adipocyte function.

Components of Dynamic Innervation Patterns

During development, the SNS establishes functional circuits with peripheral target organs. This process is distinguished into specific growth stages: 1) axonal growth initiation; 2) proximal axon extension (for long-distance growth along the vasculature); 3) distal axon extension (for short-distance growth within peripheral target organs, not bound to vasculature); 4) final target innervation and neuronal survival (formation of axonal varicosities and neuronal survival signals); and 5) growth inhibition (growth-restricting and apoptotic signals) (71, 72). Importantly, these steps in axonal growth require interaction of growth factors and their receptors to guide axonal growth toward their targets. Final target innervation, neuronal survival, and growth inhibition seem to fit well with the observed adipocyte innervation patterns and dynamic changes. Several growth factors, including the nerve growth factor (NGF), and their interaction with according receptors, e.g., tyrosine kinase A (TrkA), are involved to regulated neuronal growth and survival and have been studied in adipose tissue. NGF is expressed in target tissues, and NGF is enriched in the adipose tissue stromal vascular fraction (73) that contains endothelial cells and preadipocytes. Similarly, NGF expression is highest in undifferentiated preadipocytes and decreased in mature white adipocytes (74), in line with the lack of sympathetic varicosities close to mature white adipocytes. Conversely, TrkA is expressed in sympathetic neurons, and genetic manipulation of this neurotrophic system at the target cell or in sympathetic neurons profoundly diminishes innervation density (71).

Dynamic innervation changes in peripheral organs of adult organisms are not well studied but are notable features in WAT exposed to physiological challenges like chronic cold exposure, which can be mimicked by chronic β3-AR stimulation (63, 73, 75). Several mouse models exist with decreased adipose tissue innervation and obesity that have genetic deletions in common with a disruption of the neuro-adipocyte communication axis and highlight the importance of cross talk between sympathetic neurons and adipocytes to maintain proper innervation and body weight homeostasis [TrkA deletion in sympathetic neurons (76), impaired secretion of growth factors NGF (77), S100B (78) from adipocytes, impaired beige adipogenesis by BMP8b (79), PRDM16 (63), and sarcolipin (80) deletion].

Importantly, growth factors that are secreted by adipose tissue can also act centrally to influence adipose tissue innervation density. This was demonstrated for the adipokine leptin. Leptin’s neurotrophic function had been recognized earlier for central circuits (81, 82), but the relevance of leptin to maintain normal sympathetic innervation in iBAT and iWAT was just recently revealed. Wang et al. (29) showed that leptin signaling in the arcuate nucleus is critical to maintain proper innervation of iBAT and iWAT. Lack of leptin signaling in either POMC or AgRP neurons resulted in a strong decrease of sympathetic fibers in both iBAT and iWAT depots. Importantly, decreased hypothalamic leptin signaling (aka leptin resistance) of diet-induced obese mice causes reduced POMC and AgRP innervation of the PVN (81), as well as reduced sympathetic innervation of iBAT and iWAT and contributed to obesity by reducing thermogenesis (29). The central neurotrophic effects are mediated by brain-derived-neurotrophic-factor (BDNF) expression in PVN neurons (29) and are in line with its role to maintain POMC and AgRP innervation of the PVN (81). However, BDNF is unlikely to act anterogradely on adipose tissue innervation density, and sympathetic innervation levels of iBAT and iWAT might instead involve classic transmitter release to maintain proper stimulation levels of pre- and postganglionic sympathetic circuits.

Regulatory Components of Sympathetic Transmission and Metabolic Effects

In contrast to long-term neurotrophic effects (several days), postganglionic sympathetic activity concerns immediate (seconds) changes in norepinephrine release to adipose tissue. As we highlighted in FIGURE 1, the activity of sympathetic neurons depends on central circuits within the hypothalamus and brainstem that either directly or indirectly change the activity level of pre-and postganglionic neurons to control the release of norepinephrine. Yet, several additional layers of regulation have been recognized with significant impact on adipose tissue beiging, metabolism, and finally body weight regulation.

Norepinephrine Bioavailability

The amount of norepinephrine that is bioavailable for adipocytes depends on the norepinephrine release at axonal varicosities, but is counteracted by reuptake transporters (Slc6a2) in presynaptic membranes of the sympathetic neuron (FIGURE 2, bullet 4). Pharmacological inhibition of these transporters, e.g., by fenfluramine, are powerful antiobesity drugs, despite their cardiovascular risk (83). An emerging concept is that the harmful cardiovascular and hyperthermic effects are mediated by the CNS, while the beneficial antiobesity effects were fully preserved by peripheral actions that involved increased excitability of sympathetic neurons (84). The critical role of peripheral norepinephrine bioavailability is further supported by the norepinephrine reuptake and degradation in adipose tissue immune cells with significant effects on adipose tissue beiging, metabolism and body weight regulation (85–87). This has also relevance in humans, even though the genetic machinery to degrade norepinephrine was also enriched in mature adipocytes and is also relevant for metabolic changes with aging (88).

Adrenergic Signal Transduction

The interface for sympathetic signaling via norepinephrine are adrenergic receptors, specifically, β3-adrenergic receptors (β3-AR) that are enriched in brown and white adipocytes of mice (89–91) (FIGURE 2, bullet 5). Functional evidence comes from studies with the highly selective β3-AR agonist CL-316,243 (CL), which robustly increases energy expenditure, fat oxidation, pancreatic insulin release, and insulin-dependent and -independent glucose uptake (92, 93) and suppresses food intake. These metabolic CL effects are highly specific to β3-AR action in brown and white adipocytes (94, 95). Such data clearly link sympathetic activity in adipose tissue with whole body metabolism, even though it is still unclear how β3-AR signaling results in regulation of pancreatic insulin release or modulation of food intake. These data have led to extensive use of CL as a thermogenic agent and a surrogate to induce iBAT thermogenesis, iWAT lipolysis, and beiging (96–100).

Despite these well-accepted concepts, select β3-AR reexpression in brown adipocytes in otherwise β3-AR knockout mice is insufficient to reinstate CL effects on metabolism, food intake, and glucose homeostasis. However, β3-AR reexpression in brown and white adipocytes fully reinstated CL effects (95), suggesting that β3-AR signaling in brown adipocytes might be necessary but not sufficient to induce thermogenesis in vivo. Similarly, genetic impairment of lipolysis in brown adipocytes (considered as critical step for BAT thermogenesis) was surprisingly without effect on β3-AR-or cold-induced iBAT thermogenesis, while deletion in brown and white adipocytes abolished β3-AR-induced BAT thermogenesis and cold-induced thermogenesis in fasted animals (101, 102). Together these data support the idea that sympathetic input to brown and white adipocytes has to work in synchrony for proper thermoregulatory control.

These data are in line with earlier findings of differential sympathetic tone in BAT and WAT during select physiological challenges (38, 39), where increased metabolism is associated with iBAT and iWAT sympathetic activation (cold-exposure, MC4R activation), while decreased metabolism is associated with iWAT, but not iBAT, sympathetic activation. Taken together these data support the idea that combined sympathetic β3-AR stimulation of WAT lipolysis and BAT thermogenic pathways is required to enable iBAT thermogenesis and that differential regulation of sympathetic activity in iBAT versus iWAT is involved to control diverse physiological adaptations.

In this context it is particularly paradoxical how the sympathetic activation of iWAT lipolysis is functionally aligned with the dramatic innervation difference among white and brown adipocytes. Another peculiar phenomenon is that denervation of iBAT greatly increases iWAT beiging but lacks metabolic effects (21, 68, 103). De novo lipogenesis (DNL) via fatty acid synthase in adipocytes is tightly coupled to lipolysis, and lipolysis robustly increases DNL (99). Interestingly, genetic perturbation of DNL has profound effects on adipocyte beiging, innervation and sympathetic activity in iBAT and iWAT, even though cold-induced thermogenesis was not affected (104). Thus an overall picture emerges where adipose tissue lipid fluxes have a key role to couple iBAT and iWAT innervation and thermogenic function, even though the exact nature of this cross talk, e.g., via circulating nutrient or adipokines remains unclear.

Signaling via β3-AR has remained the focus for obesity research until today, even though it is known that genetic deletion of β3-AR is insufficient to cause obesity and only genetic deletion of all three β-adrenergic receptors (β-less mice) results in enhanced weight gain on a calorie rich high-fat diet (105). Also, the translation of β3-AR-induced thermogenesis to in humans has been difficult, even though the newly developed drug mirabegron has recently shown promising results (106). Thus reexamining the importance of other adrenergic receptors for body weight homeostasis and their interaction with β3-AR signaling is an upcoming and important future task.

Cold-exposed β-less mice or treatment with β-blockers recently revealed that adipocyte beiging occurs even in the absence of β-AR signaling and recruits beige adipocytes from muscle-like cells that contribute to thermogenesis and glucose uptake (107). It remains unclear how cold-induced sympathetic tone could induce this beiging event in the absence of β-AR signaling. Another recent finding highlighted the importance of α2-AR signaling for the regulation of adipocyte released circulating leptin levels (108, 109). Signaling via α2-AR involves G_i-coupling and is best known to counteract G_q-coupled β-AR signaling via adenylate cyclase. However, all β-AR receptors can also switch to an alternative G_i-coupling and engage an alternative MAPK signaling pathway that paradoxically results in enhanced thermogenic function (110).

Another recent study found that β2-AR are the main thermogenic receptor in human adipocytes, with comparable function to β3-AR in mouse adipocytes (111). Similarly, a recent study showed the importance of β2-AR signaling to increase sympathetic excitability and its role for thermogenesis and adipose tissue beiging (84). Future studies need to re-address tissue-specific functions of adrenergic receptor subtypes and address possible confounding effects of germline deletion in contrast to inducible deletion systems.

Adipokine Feedback and Cross Talk

Adipokines (including leptin and adiponectin) effect other hormones like glucocorticoids and insulin/glucagon levels provides feedback signals to the hypothalamus (29, 31, 48, 49, 112) and cross talk with other tissues (2, 113, 114). (FIGURE 2, bullets 7 and 8). The sympathetic regulation of adipose tissue is required for leptin gene expression and secretion and similarly leptin regulates sympathetic innervation density in iBAT and iWAT and iWAT beiging (29–31). The implementation of optogenetic and chemogenetic technologies to directly stimulate and inhibit sympathetic nerves to select adipose tissue depots could greatly facilitate our understanding of the neuro-adipocyte interface and adipokine mobilization. First studies used optogenetic activation of sympathetic nerve endings in iBAT and showed the expected increase in BAT temperature and glucose uptake (115, 116), in contrast to brown adipocyte specific β3-AR signaling, which was unable to increase thermogenesis (95). Also, select optogenetic activation of iWAT was able to induce lipolysis (117). However, effects on adipokines, iWAT beiging, and innervation changes were not investigated in this study. It should be cautioned that conditional strategies to target the peripheral nervous system require thorough characterization of marker genes to ensure proper expression profiles. For example, tyrosine hydroxylase is an excellent marker to distinguish sympathetic from parasympathetic neurons in adult mice; however, its transient expression in parasympathetic precursors during development (118) makes it unsuitable to distinguish these peripheral neurons by germline procedures (33) and will require manipulation strategies in adult animals for proper targeting. Further progress of these technologies is also a prospective avenue to further understand the importance of sensory nerves in the communication of adipocytes and CNS as suggested earlier (119–121).

Adipose Tissue Innervation and Sex Differences

The central nervous system is well acknowledged to harbor several sites and neuronal populations with pronounced sex differences and differential expression of estrogen and androgen receptors (26, 122, 123). POMC neurons, which have profound effects on adipose tissue beiging (29, 31), are also regulated by sex hormones during early life and regulate the number and innervation density of POMC neurons (124), metabolism, insulin sensitivity, and body weight in adult mice (122, 125), even though adipose tissue beiging and innervation were not directly assessed in these experiments. Similarly, sex-specific obesity phenotypes are also mediated by adipokine feedback via vagal afferents where effects on adipose tissue must be mediated indirectly by peripheral tissue cross talk (114, 126).

In line with this, sex differences are well known for adipose tissue distribution in humans with more subcutaneous, but less visceral, adipose tissue in women and increased genetic markers for beiging were found in visceral abdominal fat in women compared to men (127), even though another study using PET scan did not show differences in beige/BAT based on gender (128). These studies did not look for sympathetic innervation and the role of melanocortins in human adipose tissue, and beiging has not been confirmed so far. Furthermore, in mice sex hormones act directly on adipocytes to enhance adipose tissue beiging (129), further supporting that sex-dependent effects on adipose tissue beiging are mediated via multiple axes, including central and peripheral routes.

Summary and Outlook

Despite these clear interactions of innervation, adipose tissue function, and body weight, it is notable that the mere denervation of BAT or WAT does not result in frank obesity (68), even though adipose tissue weight increases (130). Robust states of adipose tissue beiging, do not necessarily result in measurable metabolic enhancement in mice (103, 131) or weight loss in humans (106, 132). The mixed outcomes for brown/beige adipose tissue and weight loss have reengaged the discussion if adipose tissue is a viable target for body weight control (10, 133).

In this review we highlighted the many layers that contribute to proper and dynamic regulation of adipose tissue innervation, regulation of sympathetic tone, and bioavailability of norepinephrine for adipocyte function. The provided model for the interaction of neuronal and peripheral signaling emphasizes that manipulations at any level will cause similar changes in body weight and that future studies should address the interface of neuron/adipocyte interaction in more detail. Given the historic difficulty to treat obesity safely and over the long term, targeting the peripheral interface will likely be a meaningful and safer target for pharmacological intervention.

Furthermore, we have outlined that the differential sympathetic activation of iBAT versus iWAT and the importance of acute and chronic changes in energy expenditure and beiging is not well characterized. Further progress using viral molecular-genetic tools in the peripheral nervous system will hopefully move the field forward to address these important questions.

Finally, we note that the role of sex hormones in the neuron-adipocyte interface can be integrated in the reviewed model. However, sex differences in adipose tissue distribution are not fully explained, but working toward a common model of a neuro-adipose interface to capture experimental modulations at the level of sex hormones, immune function, as well as obesity and cachexia is an important step for meaningful strategies to treat human illness.