The recognition of sensory signals from within the body (interoceptive) and from the external environment (exteroceptive), along with the integration of these cues by the central nervous system, plays a crucial role in maintaining metabolic balance. This orchestration is vital for regulating processes related to both food intake and energy expenditure. Animal model studies indicate that manipulating specific populations of neurons in the central nervous system which influence these processes can effectively modify energy balance. This body of work presents an opportunity for the development of innovative weight loss therapies for the treatment of obesity and type 2 diabetes. In this overview, we delve into the sensory cues and the neuronal populations responsible for their integration, exploring their potential in the development of weight loss treatments for obesity and type 2 diabetes. This article is the first in a series of Perspectives that report on research funded by the American Diabetes Association Pathway to Stop Diabetes program.
Anticipatory responses triggered by sensory cues associated with food intake and predatory threats recruit hypothalamic neuronal populations, which are actively involved in energy regulation.
Peripheral and central thermoregulatory neurons modulate brown adipose tissue thermogenesis, and the manipulation of thermoreceptors is associated with adiposity changes in obesity models.
The therapeutic potential of leveraging the thermogenic mediators of psychological stress–induced hyperthermia remains unexplored.
Introduction
Obesity is a leading risk factor for developing type 2 diabetes, often in conjunction with low physical activity and a poor diet (1,–,3). Severe dysfunctions in controlling food intake and energy dissipation processes are frequently associated with the progression of obesity. Substantial evidence suggests that while obesity may initially result from lifestyle factors, it can disrupt the body’s energy balance regulation due to impaired signaling in the hypothalamus (4). Therefore, understanding the cellular components controlling energy homeostasis in the hypothalamus is of key significance for the treatment of metabolic syndrome and type 2 diabetes.
The regulation of body weight occurs centrally, involving the integration of peripheral hormonal signals released from organs like the gastrointestinal tract, pancreas, and adipose tissue. The primary center for this integration is the hypothalamus, which controls food intake and energy expenditure. The list of identified peripheral appetite modulators is continuously growing and encompasses hormones such as leptin, ghrelin, cholecystokinin (CCK), peptide YY, insulin, pancreatic polypeptide, and glucagon-like peptide 1 (GLP-1). While we have gained a significant understanding of internal signals integrated by the hypothalamus, the impact of external perception on energy balance has been overlooked. Earlier work demonstrated that the loss of sensory detection modalities could impact weight gain in mice and metabolic health throughout the life span (5,6). More recently, new research shows that sensory cues are also processed in hypothalamic nuclei, and this neuronal activity affects energy balance. These new findings raise the question as to whether multiple neuronal populations are involved in the integration of internal and external signals and whether the combined influence of these factors shares synergistic or antagonistic properties in energy balance and weight gain. For example, the perception of stressful cues or episodic stress episodes might increase or reduce the influence of appetite signals on hypothalamic neurons, and these effects could alter energy homeostasis durably beyond immediate detection.
In this context, we discuss recent progress in identifying neuronal populations involved in the integration of interoceptive and exteroceptive cues and their subsequent impact on physiology. We highlight how the Pathway to Stop Diabetes–funded research has permitted the identification of neuronal targets that integrate sensory cues and alter energy metabolism as well as their potential in the development of weight loss therapies relevant to obesity and type 2 diabetes. Throughout this discussion, we provide insights into potential future pathways concerning central targets that regulate energy balance. We explore the challenges posed by existing experimental data and propose approaches to address these significant unanswered questions.
Negative-Feedback Circulatory Signals Regulating Energy Balance via Hypothalamic Neurons
Multiple circulatory factors, including adiposity, hormonal, and nutrient-related signals, help maintain stable body weight by adjusting food intake and energy expenditure in response to changes in energy status via communication with hypothalamic neurons, therefore acting as negative-feedback signals (7) (Fig. 1). The study of these various physiological cues has been key to the discovery of neuronal cell types involved in the regulation of energy balance. Among these signals, the adipose-derived hormone leptin harbors the most profound impact on appetite and energy expenditure through anorexigenic effects (8,9). The primary target of leptin action is postulated to be the hypothalamic arcuate nucleus (ARC), located in the mediobasal hypothalamus and adjacent to the base of the third ventricle and median eminence (10). Administering leptin through intracerebroventricular injection into the ARC reduces food intake and weight gain (11). Other hypothalamic areas, including the dorsomedial hypothalamic (DMH) and ventromedial hypothalamic (VMH) regions, also express leptin receptors and are directly activated by leptin, influencing energy balance in diet-induced obesity (DIO) models (12,13). Leptin receptors (LepR) are also found within the lateral hypothalamus (LH) as well as in extrahypothalamic areas such as the nucleus of the solitary tract, the parabrachial nucleus, periaqueductal gray, and dorsal raphe, among others (14).
Sensory cue detection by hypothalamic nuclei regulating energy homeostasis. The cartoon depicts the integration of food signals (green), temperature stimuli (blue), and threat perception (gray) by hypothalamic nuclei, where neuronal activity is associated with changes in feeding and/or energy expenditure processes. Figure created with BioRender.com.
Sensory cue detection by hypothalamic nuclei regulating energy homeostasis. The cartoon depicts the integration of food signals (green), temperature stimuli (blue), and threat perception (gray) by hypothalamic nuclei, where neuronal activity is associated with changes in feeding and/or energy expenditure processes. Figure created with BioRender.com.
Agouti-related peptide (AgRP) and pro-opiomelanocortin (POMC) neurons are the two most extensively studied neuronal populations responsible for regulating feeding behavior and energy expenditure (15). When fasting occurs, AgRP neurons become active, driving the desire for food and increasing food consumption. Conversely, POMC neurons are suppressed following a meal, promoting feelings of fullness and satiety. Both types of neurons coexist in the ARC and send projections to a shared set of subcortical structures, where they exert contrasting influences on food intake and various autonomic and behavioral responses influenced by energy balance. This intricate interplay governs our overall energy regulation. The ARC plays a key role in orchestrating responses from various hormones like leptin and insulin, nutrients such as glucose and free fatty acids, and input from other hypothalamic and extrahypothalamic regions (16). Insulin, secreted by pancreatic islets, also circulates in proportion to body fat and has anorexigenic effects through the modulation of insulin receptors on ARC neurons (17–21), yet it is a less powerful appetite suppressant than leptin.
Gastrointestinal hormones also play an essential role in the short-term regulation of feeding by controlling nutrient intake, absorption, and meal termination through negative feedback to the brain (22). Ghrelin is secreted by the stomach in anticipation of nutrients and stimulates appetite, whereas GLP-1, CCK, peptide YY, glucose-dependent insulinotropic polypeptide, and oxyntomodulin are secreted from the gut in response to nutrients and induce satiety (23). The sites of action of gut hormones have been evidenced both on the vagus nerve, the area postrema in the brainstem that, in turn, relays visceral information to hypothalamic neurons, as well as in direct action on hypothalamic neurons (24). For example, central GLP-1 receptors are required for the anorexigenic and weight loss effects of the GLP-1 receptor agonist liraglutide (25,26). Ghrelin receptors are also found in the hypothalamus, in regions including the ARC, VMH, and paraventricular (PVH) nuclei of the hypothalamus (27,28). Furthermore, nutrients, including glucose and fatty acids, also modulate hypothalamic neuronal activity. Recent research has demonstrated that the infusion of nutrients directly into the stomach, such as sugars, fats, and proteins, promptly reduces the activity of AgRP neurons relative to calorie ingestion (29,30).
Environmental Temperature Perception and Negative Energy Balance Feedback Pathways Controlling Thermoregulation
Sustaining a steady body temperature is crucial for staying alive, as alteration of temperature by even a few degrees could lead to serious and incapacitating consequences, such as hyperthermia or hypothermia. The detection and integration of temperature cues is associated with physiological and behavioral measures to cool down or warm up the body (31). Brown adipose tissue (BAT) thermogenesis, skeletal muscle shivering, water evaporation (sweating), and the constriction of blood vessels are the major autonomic measures involved in thermoregulation. Behavioral measures include cold or warm seeking, such as nesting or social huddling in rodents or wearing clothes and changing the room temperature in humans. The regulation of body temperature is closely intertwined with the systems that maintain energy homeostasis, as the mechanisms responsible for temperature regulation impose significant demands on the body’s resources. For instance, in the case of cold-induced thermogenesis, around 60% of the overall energy expended by mice is directed toward generating heat when exposed to ambient cold (32). To meet this increased energy requirement, mice that experience cold temperatures for prolonged periods will augment their daily food consumption, effectively doubling their intake (33). This cold-induced hyperphagia phenotype has been attributed primarily to activation of AgRP neurons in the ARC (34). Recently, the Xiphoid nucleus of the thalamus was found to contain neurons that also drive cold-induced food seeking in response to increased energy demand from the body (35), suggesting that a number of neuronal populations, in addition to AgRP neurons, are recruited by cold exposure to restore homeostasis.
Initial investigations into the regulation of body temperature have indicated that skin thermoreceptors perceive ambient temperature and convey this information to the brain via the spinal cord and midbrain (36–38). Cutaneous thermoreceptors, which are primary sensory nerve endings located in the skin, detect cold stimuli through transient receptor potential (TRP) channels such as TRPM8 receptors, specifically being gated by mild cold <26°C (39,40), and TRPV1 receptors that detect painful heat >42°C (41,42). Temperature information is sent to lateral parabrachial nucleus (LPB) and processed in the preoptic area (POA) of the hypothalamus and subsequently dispatched to peripheral tissues to initiate autonomic and behavioral adjustments (43,44). Several thermoregulatory neurons within the POA exert various negative-feedback roles to normalize temperature during deviations, including heat-sensing nitric oxide synthase 1 (NOS1), pyroglutamylated RFamide peptide (QRFP), estrogen receptor α (ERα), pituitary adenylate cyclase-activating polypeptide (PACAP), Galanin, and Bdnf (45–50) as well as cold-sensing bombesin receptor subtype-3 (BRS3) (51). Recently, another pathway recruited in cold defense was found to run directly from the LPB to the DMH without contacting the POA (52). This cold defense neurocircuit, which recruited somatostatin neurons in the LPB, induced BAT thermogenesis, muscle shivering, higher heart rate, and locomotion. Another cell type outside the POA was associated with cold-induced hyperthermia and social interaction–dependent thermogenesis and expressed prodynorphin (Pdyn) (53). Key thermoregulatory neuronal populations and their effect on energy balance are summarized in Table 1.
Key sensory perception–associated genes and the regulation of energy homeostasis
. | Role . | Expression . | Effect on appetite . | Effect on EE . | Effect on glucose homeostasis . |
---|---|---|---|---|---|
AgRP/NPY | Food and predator cues | ARC | Food/predator sensing inhibits AgRP neurons (75–77,90) | Possible role in cephalic phase response (74,75) | Neuro. stim. reduces GTT (21) |
Bdnf | Temperature cues | POA | Not reported | Neuro. stim. reduced BAT temperature (50) | Not reported |
BRS3 | Temperature cues | POA, DMH, PVH | Neuro. stim. suppresses feeding in PVH (69) | Neuro. stim. increases EE (51,69) | Mixed effects (69) |
CCK | Predator cues | DMH, PMd | Suppresses feeding (83,90) | Neuro. stim. increases EE (83) | Not reported |
CGRP | Temperature cues | DRG | No effect in Adv-CGRP KO (61) | Adv-CGRP KO increases EE (61), CGRP antibody increases EE (62) | Improves GTT in Adv-CGRP KO and db/db mice (61,62) |
ERα | Temperature cues | POA | Not reported | Not reported | Not reported |
Galanin | Temperature cues | POA (46) | Not reported | Not reported | Not reported |
LepR | Food cues | ARC, DMH, POA, PVH, LH, PM | Neuro. stim. suppresses feeding (79,80) | Neuro. stim. increases EE (79,80) | Neuro. stim. improves GTT (14) |
NOS1 | Temperature cues | POA (48) | Not reported | Not reported | Not reported |
PACAP | Temperature cues | POA | Not reported | Neuro. stim. reduced BAT temperature (50) | Not reported |
Pdyn | Temperature cues | VMH | Not reported | Neuro. stim. increases BAT temperature (53) | Not reported |
POMC | Food cues | ARC | Food sensing stimulates POMC neurons (75) | Activation drives sympathetic outflow (74) | mixed effects |
QRFP | Temperature cues | POA, LH | Not reported | Neuro. stim. reduces EE (45) | Neuro. stim. reduced glucose levels (45) |
SF1 | Predator cues | VMH | Mixed effects (98–102) | Mixed effects (98–102) | SF1-IR KO improves GTT on HFD (98) |
Somatostatin | Temperature cues | LPB | Not reported | Neuro. stim. increases EE (52) | Not reported |
TRPV1 | Temperature cues | DRG | No effect in TRPV1 KO (5) | Increases EE in TRPV1 KO (5) | Improves GTT in TRPV1 KO (5,63) |
TRPM8 | Temperature cues | DRG | Hyperphagia in TRPM8 KO (67) | No effect in TRPM8 KO (67); icilin injection increases EE (65) | Impaired GTT in TRPM8 KO (67) |
. | Role . | Expression . | Effect on appetite . | Effect on EE . | Effect on glucose homeostasis . |
---|---|---|---|---|---|
AgRP/NPY | Food and predator cues | ARC | Food/predator sensing inhibits AgRP neurons (75–77,90) | Possible role in cephalic phase response (74,75) | Neuro. stim. reduces GTT (21) |
Bdnf | Temperature cues | POA | Not reported | Neuro. stim. reduced BAT temperature (50) | Not reported |
BRS3 | Temperature cues | POA, DMH, PVH | Neuro. stim. suppresses feeding in PVH (69) | Neuro. stim. increases EE (51,69) | Mixed effects (69) |
CCK | Predator cues | DMH, PMd | Suppresses feeding (83,90) | Neuro. stim. increases EE (83) | Not reported |
CGRP | Temperature cues | DRG | No effect in Adv-CGRP KO (61) | Adv-CGRP KO increases EE (61), CGRP antibody increases EE (62) | Improves GTT in Adv-CGRP KO and db/db mice (61,62) |
ERα | Temperature cues | POA | Not reported | Not reported | Not reported |
Galanin | Temperature cues | POA (46) | Not reported | Not reported | Not reported |
LepR | Food cues | ARC, DMH, POA, PVH, LH, PM | Neuro. stim. suppresses feeding (79,80) | Neuro. stim. increases EE (79,80) | Neuro. stim. improves GTT (14) |
NOS1 | Temperature cues | POA (48) | Not reported | Not reported | Not reported |
PACAP | Temperature cues | POA | Not reported | Neuro. stim. reduced BAT temperature (50) | Not reported |
Pdyn | Temperature cues | VMH | Not reported | Neuro. stim. increases BAT temperature (53) | Not reported |
POMC | Food cues | ARC | Food sensing stimulates POMC neurons (75) | Activation drives sympathetic outflow (74) | mixed effects |
QRFP | Temperature cues | POA, LH | Not reported | Neuro. stim. reduces EE (45) | Neuro. stim. reduced glucose levels (45) |
SF1 | Predator cues | VMH | Mixed effects (98–102) | Mixed effects (98–102) | SF1-IR KO improves GTT on HFD (98) |
Somatostatin | Temperature cues | LPB | Not reported | Neuro. stim. increases EE (52) | Not reported |
TRPV1 | Temperature cues | DRG | No effect in TRPV1 KO (5) | Increases EE in TRPV1 KO (5) | Improves GTT in TRPV1 KO (5,63) |
TRPM8 | Temperature cues | DRG | Hyperphagia in TRPM8 KO (67) | No effect in TRPM8 KO (67); icilin injection increases EE (65) | Impaired GTT in TRPM8 KO (67) |
ARC, arcuate nucleus; BAT, brown adipose tissue; DMH, dorsomedial hypothalamus; DRG, dorsal root ganglion; EE, energy expenditure; GTT, glucose tolerance test; HFD, high-fat diet; KO, knockout; LH, lateral hypothalamus; LPB, lateral parabrachial nucleus; Neuro. stim., neuronal stimulation; PMd, dorsal premammillary nucleus; POA, preoptic area; PVH, paraventricular hypothalamus; VMH, ventromedial hypothalamus.
In humans, multiple studies have explored the health benefits of cold-induced thermogenesis. Repeated cold exposure has been associated with successful BAT thermogenic activity in lean and healthy subjects (54,–,56). Short exposure to cold (2 h at 15.5°C) in lean people with large BAT depots increases whole-body energy expenditure through lipid metabolism and potential recruitment of the thermogenic subcutaneous WAT (57). After 10 days of cold acclimatation in human subjects, the cold environment was judged to feel warmer and self-reported shivering decreased during cold exposure, allowing BAT thermogenesis (54). The variable effects of cold exposure associated with genetics, body composition, and individual tolerance to cold may discourage some individuals from adhering to this practice consistently, even though beneficial metabolic effects have been observed in obesity and type 2 diabetes (57,–,59). The genetic aspects associated with efficient cold thermogenesis are poorly known, and recent evidence indicates that they play a key role in this process. Remarkably, a mutation in α-actinin-3, found in colder European climates, diminishes muscle shivering capacity and was associated with the ability to complete a cold-water immersion test (14°C) for 120 min (60). Dissecting the regulatory components of cold-induced thermogenesis holds the potential to unlock the health benefits of this method and diminish the adverse effects.
Targeting the activity of peripheral and central thermoregulatory neurons for the control of feeding and autonomic energy dissipation processes might open avenues for weight loss therapies. Our data indicate that loss of heat-sensing calcitonin gene–related peptide (CGRP)–expressing sensory neurons, which constitute a subset of TRPV1-positive sensory neurons, was associated with higher cold-induced BAT thermogenesis, increased energy expenditure, and resistance to weight gain upon high-fat diet feeding (61). This research showed that mice lacking CGRP neurons had enhanced cold sensitivity at ambient temperature, resulting in higher BAT lipid utilization. Monoclonal antibody therapy against CGRP also reduced weight gain and improved glucose tolerance and insulin sensitivity in diabetic db/db mice (62). Due to its large size, this antibody does not cross the blood-brain barrier and allows the selective manipulation of circulating CGRP. In accordance with a beneficial health effect of reduced TRPV1/CGRP signaling in the peripheral nervous system, we also observed that genetic deletion of TRPV1 resulted in higher energy expenditure and increased life span (5) in addition to protection against DIO (63). In accordance with our findings, pharmacological blockade of TRPV1 elicited long-lasting hyperthermia, which is a consequence of high thermogenesis (36,64). Additionally, targeting TRPM8 receptors has also yielded interesting results in treating obesity. Injection of icilin, a cooling agent that stimulates the peripheral cold sensor TRPM8, drove weight loss in DIO mice (65). Conversely, pharmacological inhibition of TRPM8 provoked hypothermia (37,66), and TRPM8 knockout (KO) mice were hypometabolic and gained more weight on high-fat feeding than wild-type controls (67). The targeting of central neurons that regulate cold defense also suggests promising avenues for obesity prevention therapies. For example, pharmacological targeting of BRS3, a G protein–coupled receptor implicated in thermoregulation, has demonstrated encouraging results on body weight, feeding, and glucose homeostasis (68). BRS3 labels cold-activated neurons found within the POA, DMH, and PVH, and their stimulation increases thermogenesis and heart rate while suppressing food intake (in the PVH) (51,69). BRS3 KO mice develop obesity through increased feeding and reduced energy expenditure (70).
Feedforward Mechanisms Affecting Energy Homeostasis: Food-Associated Sensory Cues
There is good evidence for the existence of anticipatory processes, or “feedforward” signals, which prepare the body for changes in energy balance in response to sensory cues associated with food intake. The cephalic phase response is triggered by the detection of food-related perception and includes the release of insulin and other hormones to prepare for digestion and nutrient absorption (71). These mechanisms help the body adapt to expected changes in energy status before they occur. Remarkably, mice exposed to odors of familiar foods during fasting showed adipose tissue lipolysis and liver postprandial priming through the activation of the central melanocortin pathway (72,73). These food perception cues might be key contributors to engaging postprandial thermogenesis to efficiently process the incoming nutrients through increased metabolic rate in anticipation of the energy required for digestion. In support of this anticipatory response, several reports have observed changes in ARC neuronal activity induced by the sole detection of food and preceding its consumption. In particular, food perception without its ingestion was enough to alter the activity state of AgRP and POMC neurons in the ARC (74–76). Using fiber photometry recordings of freely behaving mice, Chen and colleagues (74) showed that fasting drove high AgRP neuronal activity, while suppressing POMC neuronal GCaMP calcium response. When caged food was presented, therefore allowing food perception without its consumption, a rapid inhibition of AgRP neurons and simultaneous activation of POMC neurons was measured (74). The activation of POMC by sensory perception of food cues is likely to be a crucial component of anticipatory processes associated with food digestion, as POMC stimulation triggered sympathetic nerve activity and liver endoplasmic reticulum remodeling (73). High AgRP neuronal activity during fasting is rapidly inhibited by the presence of food, consistent with a negative valence signal (75). To date, the functional significance of the preconsummatory inhibition of AgRP neuron activity remains unclear, and several hypotheses have been put forward to explain this phenomenon (77). This inhibition could serve as an anticipatory signal for an impending satiety response, preventing excessive food intake. It may aid in the transition from food-seeking behavior to food consumption and may promote the cephalic phase response.
Interestingly, the palatability of the food correlates with the extent of the response on the AgRP/POMC system, indicating that ARC neurons integrate the energetic value of the food in response to sensory stimuli. This most likely allows the neurons to receive real-time information about the availability and desirability of the food. It is important to note that the effect on AgRP/POMC neurons is reversed if the sensed food is not actually eaten and the animals remain fasted. Therefore, it is likely that AgRP and POMC neurons play a critical role in food-seeking and anticipatory processes associated with priming the organism for digestive processes.
Other neuronal players have been implicated in this anticipatory response, including neurons in the DMH. Garfield et al. (78) demonstrated that GABAergic neurons expressing LepR in the ventral DMH send projections to arcuate AgRP neurons, and the DMH-LepR→ARC-AgRP circuit inhibited feeding. DMH-LepR neurons exhibit rapid activation upon food presentation in a manner reflecting the nutritional value of the food, whereas AgRP neurons are inhibited by food presentation.
The preconsummatory inhibition of AgRP neuron activity could also act as a negative learning signal, associating food items with hunger-relieving sensations (77). Recently, a study supported this hypothesis, as LepR-expressing neurons from the DMH showed high activation in food-restricted mice that were trained on a two-alternative forced-choice task (79). Excitatory input to DMH-LepR neurons was shown to originate from the LH, which contains neurons expressing vesicular glutamate transporter 2 (Vglut2). Disruption of the LH-Vglut2→DMH-LepR→ARC-AgRP neurocircuit led to dysfunctional AgRP inhibition in response to food cues and impaired learning of a sensory cue–initiated food acquisition task.
The experience of eating is associated with a behavioral sequence, including an appetitive phase consisting of seeking behavior followed by a consummatory phase, where the animal localizes the food source and manipulates the food through biting and chewing, and finally meal termination and digestion (80). Due to the distinct characteristics of seeking and consummatory behaviors in terms of motivational state and behavioral decisions, separate populations of neurons might be recruited for the control of each behavior. Recently, another population of LepR-expressing neurons found in the LH was found to influence both seeking and consummatory behaviors (81). Interestingly, this new study elegantly leveraged microendoscopic imaging analysis of GCaMP activity during the seeking and consummatory feeding phases to identify two distinct subpopulations of LH-LepR neurons, the neuronal activity of which was associated with either food seeking or consumption. These subpopulations are sequentially activated upon behavioral change. Optogenetic activation of LH-LepR neurons drove seeking and consummatory behaviors, whereas their inhibition decreased consummatory behaviors. Neuropeptide Y (NPY) appeared to play a role in tuning LH-LepR neuronal activity, but the source of NPY as well as its biological involvement in this process is not well understood. NPY has been shown to be a crucial neurotransmitter to sustain hunger after AgRP inactivation (82), and as AgRP/NPY neurons innervate the LH, it is plausible that during fasting, elevated levels of NPY resulting from increased activity in AgRP/NPY neurons relieve the continuous inhibition on LH-LepR neurons. This enables LH-LepR neurons to effectively generate the necessary appetite stimulation in response to various cues associated with food. In contrast, a fed state would be associated with reduced neural activity of AgRP/NPY, creating a constant inhibitory mechanism. This mechanism would therefore prevent LH-LepR neurons from being triggered by cues related to food. Taken together, these studies indicate that LH neurons actively participate in food seeking and consumption, and DMH and ARC nuclei regulate anticipatory responses associated with food perception and prime the organism for the ingestion of nutrients. Key food-sensing neuronal populations and their effect on energy balance are summarized in Table 1.
Feedforward Mechanisms Affecting Energy Homeostasis: Predatory Cue Sensing
In addition to facilitating the identification and retrieval of food, sensory perception also exerts a significant influence on guiding behavioral choices in the presence of potential threats as well as adaptive metabolic responses ensuring survival. Such sensing is primarily driven by the olfactory system in smaller prey animals like rodents, whereas visual information plays a dominant role in hazard identification in humans. Appropriate adaptive adjustments in both behavior and bodily functions include suppression of food intake and mobilization of energy reserves to facilitate decisions related to confronting or escaping from danger in fight-or-flight situations. A key metabolic output associated with the perception of stressors in mammals is a rapid increase in body temperature, referred to as psychological stress-induced hyperthermia (PSH). This rise in temperature likely serves to enhance both physical and neural performance by raising the temperature of muscles and the central nervous system by a few degrees Celsius. PSH represents an automatic stress reaction that activates the sympathetic nervous system, triggering a series of physiological responses and resulting in increased energy expenditure (83). Blockade of β3-adrenoreceptors efficiently reduces PSH, whereas cyclooxygenase inhibitors, which are used to decrease inflammation-induced fever, are ineffective (84,–,86). PSH has been observed in the context of psychological stress under the perception of danger, discomfort, or pain and upon central delivery of corticotropin-releasing hormone (CRH), which is secreted by PVH neurons during the stress response (84,87,88).
The neuronal basis of PSH appears to lie within the DMH. Perception of the fear-inducing odor 2,4,5-trimethylthiazole (mT), a synthetic compound related to the red fox predator scent 2,4,5-trimethylthiazoline (TMT), is associated with cFos upregulation in the DMH in mice compared with control scent (89). Similarly, PSH resulting from psychological stress induced by social defeat stress in rats was suppressed by inactivation of DMH neurons using muscimol microinjections (88). We recently demonstrated that CCK neurons in the DMH are activated by predator smell, with stronger responses observed in female mice (83). The stimulation of DMH-CCK neurons promotes energy expenditure through increased BAT thermogenesis, whereas their neuronal silencing reduces predator-induced energy expenditure in the presence of predator odor.
Another major output of predator detection is hypophagia, recognized as a shared characteristic among numerous stress models. The neural basis for predator-induced hypophagia was recently shown to recruit CCK neurons found in the dorsal premammilary nucleus, which has activity that suppresses AgRP neurons (90). Remarkably, we also observed that DMH-CCK neurons contribute to stress-induced hypophagia, with their activation inhibiting feeding and their inhibition restoring normal food intake despite the presence of a predator odor (83). Interestingly, physical restraint selectively induced anorexigenic POMC neuron activity in ARC to suppress feeding, whereas other stressors, such as TMT, had no effect on POMC neuronal activation (91). This finding suggests that the integration of predator odors diverges from other types of stressors in terms of metabolic outcome and neuronal targets recruited, highlighting the complexity of stressful cue integration.
Other stress-dependent behaviors have been attributed to PVH-CRH neurons following footshock stress, including grooming, rearing, or walking (92). CRH neurons are also key to mediating the endocrine aspect of the stress response (92). The stimulation of CRH neurons controls the release of cortisol in humans and its rodent analog, corticosterone (CORT), through hypothalamic-pituitary-adrenal (HPA) axis activity initiated by release of CRH from the PVH into the pituitary portal (93). Stimulation of DMH-CCK neurons appears to recapitulate both stress-induced hypophagia and PSH but does not lead to endocrine release of CORT, hinting that multiple cellular effectors participate in the generation of complex behavioral and physiological aspects of the stress response.
Interestingly, many murine predator odors associated with defensive behavior, such as snake-, rat-, or hawk-related odors, specifically activate the dorsomedial ventromedial nucleus of the hypothalamus (VMHdm) (94). Steroidogenic factor 1–expressing (SF1+) neurons are a unique cellular population expressed throughout the VMH and are involved in defensive behaviors associated with various predator odors (95,–,97). Evidence suggests that subgroups of these neurons also regulate energy balance upon humoral cues, but whether this regulation is controlled by predator odors is unknown. Deletion of LepR in SF1 neurons in the VMH decreased energy expenditure and led to increased weight gain (12), whereas deletion of IR in SF1 neurons was protective against DIO through improved glucose homeostasis (98). Additionally, ERα-expressing neurons, colocalizing with SF1 in the ventrolateral VMH (VMHvl), stimulate energy expenditure by engaging locomotor activity selectively in female mice (99,100). Conversely, genetic deletion of ERα in VMH in SF1-Cre females was associated with hypometabolism, hyperphagia, reduced energy expenditure, and increased adiposity (101), confirming previous observations from ERα RNA interference in the VMH (102). Further investigations are required to explore the role of VMH neurons in promoting PSH upon predator cue sensing.
Considering the stress response's capacity to control both appetite and energy expenditure, exploring the modulation of predator-triggered neurons for the development of weight loss strategies could be an interesting approach in obesity and diabetes. This is particularly relevant as different neuronal targets appear to induce endocrine and physiological outputs of stress perception, with therapies selectively targeting PSH and hypophagia, without detrimental glucocorticoid release. Chronic elevation of cortisol resulting from HPA overactivity is undesirable in obesity treatment (103). Prolonged elevation of CORT levels in rodents is associated with weight gain and increased adiposity, through the alteration of feeding behavior leading to hyperphagia (104), without affecting energy expenditure (105,–,107). It is important to recognize that in many models of chronic stress, prolonged hyperphagia and behaviors resembling binge eating in rodent models have been observed (108,–,110). The desire for high-calorie meals is likely elevated due to increased levels of glucocorticoids and insulin (104), and increased hypothalamic expression of AgRP and NPY mRNA has been observed (111). However, the observation that chronic exposure to the fox-related predator scent mT mitigated DIO in mice, despite driving CORT release in mice, demonstrates an adaptive response to chronic exposure to predatory scents (89). In this model, neither food intake nor locomotion was altered over time. Significantly, the effectiveness of stress in inducing PSH has been observed in humans, indicating the potential applicability of this pathway in translational research. An immediate psychological challenge experienced under conditions of neutral temperature has been linked to a rise in supraclavicular BAT temperature (112). Better understanding of the various cellular components involved in the central processing of predator cues as well as better characterization of their metabolic outputs is key to leveraging the negative energy balance pathways associated with stress perception for weight loss strategies. Key stress-sensing neuronal populations and their outpout on energy homeostasis are summarized in Table 1.
Conclusion and Perspectives
In conclusion, as we delve into the molecular complexity of hypothalamic neuronal populations, several important findings shedding light on the regulation of energy balance have emerged. Recent advances in our understanding of food-associated sensory processing have highlighted the crucial involvement of neuronal circuits in translating external cues into behavioral and physiological responses. Neurons within the DMH and LH have been implicated in the modulation of food-seeking and consumption behaviors, shedding light on the intricate orchestration of energy homeostasis through sensory perception. Predator detection's link to energy balance, exemplified by the phenomenon of PSH, has unveiled the role of specific hypothalamic neurons, such as those expressing CCK. These neurons not only impact thermogenesis but also contribute to stress-induced hypophagia, showcasing their potential as therapeutic targets to modulate energy expenditure and appetite. However, despite these remarkable insights, challenges remain. Molecular redundancy within the brain complicates efforts to precisely target specific neurons for therapeutic interventions. The advent of single-cell RNA sequencing has provided a powerful tool to distinguish molecular signatures and uncover distinct neuronal populations. This technique holds promise in identifying unique cellular markers for improved precision in targeting unique cell populations but also showcases the heterogeneity of hypothalamic cell types, as recently highlighted (113). For example, different subsets of POMC- and AgRP-expressing neurons exhibit distinct transcriptional responses to changing energy status and might play different metabolic roles (114,– 116).
In the realm of obesity treatment, the convergence of research on hormonal cues and sensory perception pathways presents exciting opportunities for novel therapeutic strategies. By harnessing the knowledge gained from these diverse avenues, researchers can develop interventions that leverage the intricate neural circuits controlling energy balance, offering new opportunities for effective treatments for obesity and metabolic disorders. As the field continues to advance, a multidisciplinary approach that combines molecular biology, neuroscience, and clinical research will be essential to unlock the full potential of these discoveries. An important challenge will be to harness the beneficial effects of these molecular targets on metabolism without negative consequences on emotional and physical well-being (117).
This article is featured in a podcast available at diabetesjournals.org/diabetes/pages/diabetesbio.
Article Information
About the Pathway to Stop Diabetes Program. The Pathway to Stop Diabetes program from the American Diabetes Association aims to create the conditions that foster scientific breakthroughs in diabetes research. Talented early-career scientists who demonstrate exceptional innovation, creativity, and productivity receive 5–7 years of funding to explore new ideas without traditional project constraints. Pathway awardees are also paired with world-renowned diabetes scientists who offer mentorship, as well as scientific and professional guidance, throughout the duration of their grant. More information on the Pathway to Stop Diabetes program can be found at https://diabetes.org/research/pathway.
Acknowledgments. During the course of preparing this work, the author(s) used Grammarly for the purpose of editing and correcting grammar, spelling, and punctuation. Following the use of this tool/service, the author formally reviewed the content for its accuracy and edited it as necessary. The author takes full responsibility for all the content of this publication.
Funding. Research in the Riera laboratory is supported by grants from the American Diabetes Association Pathway to Stop Diabetes grant 1-15-INI-12 (C.E.R.) and the Klingenstein-Simons foundation (C.E.R.).
Duality of Interest. No potential conflicts of interest relevant to this article were reported.