The essential role of the brain in maintaining energy homeostasis has motivated the drive to define the neural circuitry that integrates external and internal stimuli to enact appropriate and consequential metabolic and behavioral responses. The hypothalamus has received significant attention in this regard given its ability to influence feeding behavior, yet organisms rely on a much broader diversity and distribution of neuronal networks to regulate both energy intake and expenditure. Because energy balance is a fundamental determinant of survival and success of an organism, it is not surprising that emerging data connect circuits controlling feeding and energy balance with higher brain functions and degenerative processes. In this review, we will highlight both classically defined and emerging aspects of brain control of energy homeostasis.
A fundamental requirement for an organism’s survival is its ability to maintain a homeostatic internal metabolic environment. Times of scarcity require sufficient energy storage from times of repletion to utilize for both basic physiological functions as well as additional energy-expensive demands such as foraging for food, flight from predators, and reproduction. This occurs on varying scales, from a daily rhythm of fasting while sleeping and feeding while awake to a longer-term cycle of famine that typified hunter-gatherer societies and especially early post-agricultural humans. As such, malfunction of energy storage machinery confers a distinct disadvantage, especially when resources are limiting. Humans’ evolutionary history has featured almost exclusively times with a dearth of these resources, with only a portion of the population relatively recently finding themselves with excess caloric availability. Coupled with modern conveniences and an associated sedentary lifestyle, these realities have led to an unabated increase in obesity, metabolic disease, and type II diabetes. The costs attributed to this increase are devastating, with an estimated hundreds of billions spent each year on medical treatment alone, alongside a reduction in work productivity and quality of life that is more difficult to quantify. While general environmental measures such as diet and exercise are often capable of alleviating some of these negative consequences, there exists an undeniable genetic component seemingly tuned in humans toward efficient and excessive energy storage that makes this process much more challenging.
Many hypotheses have been put forth to explain the origins of this efficiency. Over 50 years ago, the “thrifty genotype” hypothesis suggested that alleles conducive to energy storage were selected for during the evolutionary history of humans to ensure the ability to survive until reproductive age in an energy-scarce environment—a genetic complement that, unfortunately, becomes detrimental in obesogenic environments. Thirty years later, an alternative stance, dubbed the “thrifty phenotype” hypothesis, placed the cause of humans’ susceptibility to obesity and type II diabetes on selective pressure due to the nutrient-poor fetal and early postnatal environment rather than that of later development. More recently, a further iteration, the so-called “drifty genotype” hypothesis, offers that these genes are, in actuality, under little selection pressure and came about due to the random nature of genetic drift. Of course, it has been further suggested that none of these models fully captures what is likely a very complex reality and that different migrating populations presumably faced varying selection environments, helping to explain the inequality in predisposition to obesity and its related metabolic disorders across groups.
Though the root cause of the mechanisms responsible for efficient energy storage may remain contested, the central role of the brain in regulating energy balance is undisputed. The act of eating is an everyday part of the human experience and represents an obvious component of energy homeostasis. Indeed, through this process, the body acquires the building blocks required for internal biochemical reactions, without which survival would cease in a matter of weeks. As such, investigation into the neural basis of the promotion or cessation of this behavior has been a major focus of metabolic research. The melanocortin system of the hypothalamus has emerged at the forefront of this work with several seminal findings showing the potent ability of its constituent cells to induce both hypophagia and hyperphagia. Though this system has been heavily studied, new paradigms continue to emerge, including that the strict orexigenic or anorexigenic definitions of populations of its constituent cells are actually more flexible. Additionally, an expanding body of data, building upon initial observations more than 30 years old, has shown that many other sets of coordinating neural circuitry are required to keep metabolic regulation in check, including extrahypothalamic networks.
To maintain energy homeostasis, mechanisms of food intake must be balanced by those regulating energy expenditure, including thermoregulation, basal metabolism, and physical activity. Indeed, appropriate recommended caloric requirements can only be calculated by utilizing such information. These complex realities highlight the exquisite control that needs to occur to maintain a reasonably homeostatic metabolic environment, especially at the neural level, and ongoing studies investigating the neural control of metabolic state have begun to unravel the mechanisms responsible for such critical processes. This work has proven essential in moving toward understanding the full scope of how the brain controls energy state and allowing for the ability to maximally target this regulation toward positive outcomes.
The hypothalamus consists of multiple distinct nuclei responsible for a host of functions through their secretion of neuroendocrine molecules—including sleep and arousal, fatigue, thermoregulation, hunger, and thirst—underscoring its critical role in the neural maintenance of energy balance. Indeed, early studies pinpointed the hypothalamus as a feeding center. For instance, work over 60 years old found that lesions in the ventromedial hypothalamus (VMH) of the rat induced significantly increased feeding whereas those in the ventrolateral hypothalamus (VLH) led to opposite feeding behavior and malnutrition, and subsequent research noted the importance of several other hypothalamic nuclei.
Overview of selected energy-balance-regulating neural circuitry involving hypothalamic neurons. Arrows represent activating inputs whereas blunt-ended lines represent inhibitory inputs; red represents connections that promote positive energy balance, whereas black represents connections that promote negative energy balance. 3V, third ventricle; AgRP, agouti-related protein; AP, area postrema; ARC, arcuate nucleus; CB1R, cannabinoid receptor 1; DMH, dorsomedial hypothalamus; GHSR, growth hormone secretagogue receptor; LepR, leptin receptor; LHA, lateral hypothalamic area; Nacc, nucleus accumbens; NPY,neuropeptide Y; NTS, nucleus tractus solitarius; PAGvl/DR, ventrolateral periaqueductal gray and dorsal raphe complex; PBN, parabrachial nucleus; PFC, prefrontal cortex; POMC, proopiomelanocortin; PVN, paraventricular nucleus; PVT, paraventricular thalamus; RPa, raphe pallidus; VMH, ventromedial hypothalamus; VTA, ventral tegmental area.
Studies determining the specific cellular basis of appetite in the hypothalamus have heavily implicated the melanocortin system in its arcuate nucleus (ARC) as a key regulator of food intake. The notion that an interplay between hypothalamic neuropeptide Y (NPY) and proopiomelanocortin (POMC) plays a role in feeding regulation was put forth in 1992. This was confirmed by the subsequent discovery of leptin in 1994 and the role of the melanocortin receptor 4 system in feeding regulation in 1997. Of particular convincing nature were subsequent ablation studies showing that loss of neurons expressing neuropeptide Y (NPY) and agouti-related protein (AgRP) induced hypophagia, weight loss, and starvation, whereas loss of a separate population expressing POMC and cocaine- and amphetamine-regulated transcript (CART) promoted the converse. Importantly, further studies of acute and reversible modulation of AgRP/NPY and POMC neurons have largely corroborated the orexigenic nature of AgRP/NPY neurons and the anorexigenic nature of POMC neurons.
Study of the molecular regulation of these two groups of neurons has shed light on their control of energy homeostasis, including through their interaction with each other. The ARC is sensitive to circulating signals pertaining to energy status through its proximity to third ventricle and modified, less restricted blood-brain barrier. Both AgRP/NPY and POMC neurons express receptors for and are targeted by the hormones leptin, a satiety signal originating from adipose tissue, and ghrelin, a hunger signal emanating from the gastrointestinal tract. Leptin inhibits activity of AgRP/NPY neurons and stimulate activity of POMC neurons, leading to the decreased feeding and increased energy expenditure via the stimulation of the melanocortin 3 and 4 receptors (MC3R and MC4R) by alpha-melanocyte stimulating hormone (α-MSH), a cleaved product of POMC. Conversely, ghrelin activates AgRP/NPY cells and inhibits POMC cells, allowing for a greater release of AgRP, a potent MC3R and MC4R antagonist, stimulating appetite. Furthermore, AgRP/NPY neurons are capable of inhibiting POMC neurons themselves through their release of NPY as well as γ-aminobutyric acid (GABA).
A recent study, though, has shown that there may be at least some flexibility within these cell types that defies such strict orexigenic/anorexigenic definitions. In this work, activation of POMC neurons via the stimulation of cannabinoid receptor 1 and subsequent release of β-endorphin actually promoted—rather than suppressed—feeding and may be the basis of cannabinoid-induced hyperphagia. Thus, though the importance of the melanocortin system is critical in the regulation of feeding behavior, the diverse means through which it is able to accomplish this control are still being understood.
One important means through which neurons in the ARC exert their effects is through relay points in other areas of the hypothalamus. AgRP/NPY neurons maintain projections to both the lateral hypothalamic area (LHA) as well as the paraventricular nucleus (PVN), and optogenetic activation of ARC > LHA and ARC > PVN projections is sufficient to induce feeding comparable to activation of AgRP/NPY neurons themselves. POMC neurons also densely innervate the PVN, as well as the LHA, the VMH, and the dorsomedial hypothalamus (DMH), among other regions. Disruption of these connections, too, modulates energy balance. For example, surgically eliminating ARC > PVN connections leads to an obese phenotype, which was proposed to be through regulation of melanocortin receptors in the PVN. Consistent with this notion, more recent work found that simultaneous optogenetic activation of AgRP/NPY neurons and MC4R-expressing PVN neurons strongly attenuates the increased food intake of activation of AgRP/NPY neurons alone, regulation that likely is extended through the action of neurons receiving input from the PVN and projecting outside of the hypothalamus to the parabrachial nucleus (PBN).
