Elsevier

Physiology & Behavior

Volume 89, Issue 4, 30 November 2006, Pages 486-489
Physiology & Behavior

Metabolic sensing neurons and the control of energy homeostasis

https://doi.org/10.1016/j.physbeh.2006.07.003Get rights and content

Abstract

The brain and periphery carry on a constant conversation; the periphery informs the brain about its metabolic needs and the brain provides for these needs through its control of somatomotor, autonomic and neurohumoral pathways involved in energy intake, expenditure and storage. Metabolic sensing neurons are the integrators of a variety of metabolic, humoral and neural inputs from the periphery. Such neurons, originally called “glucosensing”, also respond to fatty acids, hormones and metabolites from the periphery. They are integrated within neural pathways involved in the regulation of energy homeostasis. Unlike most neurons, they utilize glucose and other metabolites as signaling molecules to regulate their membrane potential and firing rate. For glucosensing neurons, glucokinase acts as the rate-limiting step in glucosensing while the pathways that mediate responses to metabolites like lactate, ketone bodies and fatty acids are less well characterized. Many metabolic sensing neurons also respond to insulin and leptin and other peripheral hormones and receive neural inputs from peripheral organs. Each set of afferent signals arrives with different temporal profiles and by different routes and these inputs are summated at the level of the membrane potential to produce a given neural firing pattern. In some obese individuals, the relative sensitivity of metabolic sensing neurons to various peripheral inputs is genetically reduced. This may provide one mechanism underlying their propensity to become obese when exposed to diets high in fat and caloric density. Thus, metabolic sensing neurons may provide a potential therapeutic target for the treatment of obesity.

Introduction

More than 50 years ago Jean Mayer predicted that arterio-venous differences in blood glucose levels were sensed by “glucoreceptors” localized in the hypothalamus as a means of controlling food intake [29]. This idea was stimulated by the Anand and Brobeck concept of lateral hypothalamic (LH) feeding and ventromedial hypothalamic (VMH) satiety centers [2]. Mayer further postulated that “the passage of potassium ions into glucoreceptor cells along with the glucose phosphate represents the point at which effective glucose level is translated into an electric or neural mechanism”. The amazing prescience of this prediction is illustrated by the fact that it was fully 20 years before Oomura [38], Anand [3] and their colleagues would actually identify neurons in the LH and VMH which used glucose as a signaling molecule to regulate their firing rates. It took an additional 22 years before such neurons were shown to utilize the K+ channel that Mayer had predicted was required to regulate glucosensing [30]. Another 10 years passed before this channel was shown to be an ATP-sensitive K+ (KATP) channel similar to the one in the pancreatic β-cell which is the final common pathway for glucose-stimulated insulin release [42]. Today, we know quite a lot about the neurons which sense and respond to ambient glucose levels. We have learned that many of these neurons also respond to a variety of metabolites and hormones from the periphery which informs them as to the metabolic status of the body [31], [47], [55]. For this reason, we have called them “metabolic sensing neurons” [19], [20], [24]. This review will describe what we currently know of the mechanisms by which these neurons sense metabolic and hormonal signals and about the ways in which this might be linked to the physiological control of food intake and energy homeostasis.

The concept of “centers” controlling satiety and hunger was a useful initial construct which focused attention on areas of the hypothalamus involved in the regulation of ingestive behavior [2], [50]. However, the brain is not organized in centers. Also, ingestion is only one aspect of a much more complex pattern of behavioral, metabolic and physiological responses involved in the regulation of energy intake, expenditure and storage, i.e. energy homeostasis. The control of energy homeostasis occurs within a distributed network of specialized metabolic sensing neurons arrayed throughout the brain and periphery. Neural afferents from the hepatic portal vein, gut, carotid sinus and other peripheral sensors travel via the vagus and sympathetic nervous system to the nucleus tractus solitarius (NTS) in the medulla and are then relayed to a variety of other brainstem, hypothalamic, limbic and neocortical areas [1], [25]. These are joined by neural inputs from somatic afferents mediating sight, sound, taste, pain and touch and these incoming signals are integrated within clusters of neurons which are also responsive to metabolic and hormonal inputs from the periphery. Unlike most neurons which utilize glucose and other metabolites to fuel the metabolic demands associated with their activity, metabolic sensing neurons use these fuels as signals to alter their activity. This was first demonstrated to be the case for glucose where increasing glucose levels led to increased activity in some neurons (glucose excited; GE) and decreases in others (glucose inhibited; GI) [3], [38]. These neurons are arrayed in clusters throughout the hypothalamus and brainstem and often co-exist in the same locations where changes in glucose levels have opposing actions on them [3], [17], [33], [38], [39]. For example, the arcuate nucleus (ARC) contains neuropeptide Y (NPY) GI [18], [34] and proopiomelanocortin (POMC) GE neurons [14], the LH contains orexin GI neurons [5] and the NTS contains both GE and GI neurons [9], [33], [59]. Up to 20–30% of neurons in such areas have glucosensing properties [15], [16], [46]. Brain areas containing glucosensing neurons are focal points where hormonal, metabolic and neural signals from the periphery converge, are integrated and then passed onward to effector areas such as the hypothalamic paraventricular nucleus and LH through which neurohumoral, autonomic and somatomotor effector systems involved in energy homeostasis are activated.

Although Mayer originally postulated that such glucosensing neurons played an integral role in the initiation and termination of meals [29], we still do not know with certainty that glucose itself is a primary regulator of hunger and satiety. There is no question that severe glucoprivation can stimulate feeding [45] and that high concentrations of glucose placed in the brain can terminate feeding [4], [12], [51], [52]. However, it is uncertain whether any of these findings implicate glucose as a primary mediator of meal to meal intake under physiological conditions. Brain glucose levels are usually only 20–30% of blood levels [10], [44] and it is certainly possible that glucosensing neurons might play some role in the regulation of ingestion since they are clearly capable of detecting and responding to the very small decrements in blood glucose levels which precede some meals in both humans and rodents [6], [7], [15], [47], [56]. However, when blood glucose levels are lowered in a stepwise fashion in humans, they report hunger only at levels slightly higher than those associated with impaired cognitive function [32]. In fact, no one has ever demonstrated that meal initiation or termination can be manipulated by altering brain glucose levels within the limits found during normal ingestion.

While their role in ingestive behavior is still in question, our understanding of the mechanisms by which glucosensing neurons respond to changes in ambient glucose levels is expanding. It appears that GE neurons sense glucose by mechanisms analogous to those operant in pancreatic β-cells. In a majority of these neurons, a high affinity hexokinase, glucokinase (GK) appears to be the primary regulator of glucosensing as it is in the β-cell [11], [15], [25], [26], [28], [57]. In non-glucosensing neurons, glucose is transported inward from the extracellular space through a high affinity transporter (GLUT3) and then glycolysis is initiated by a high affinity hexokinase (hexokinase I) [53], [54]. Glycolysis initiates a sequence of steps leading to the generation of ATP which inactivates (closes) the KATP channel. This promotes membrane depolarization, influx of calcium and propagation of an action potential [40], [42]. The mechanism by which glucose inhibits activity in GI neurons is less clear although it does appear to depend on GK, the formation of ATP and may involve a non-specific chloride channel [11], [15], [16], [46]. Since the GLUT3 transporter and hexokinase I are saturated at physiologic glucose levels and since hexokinase I is inhibited by its product, glucose-6 phosphate [53], neither can be regulators of the ATP production required to alter GE or GI neuronal activity. This role appears to be played by glucokinase (GK) in ∼ 65% of GE and ∼ 45% of GI neurons [16] as it is in the pancreatic β- and α-cells, respectively [13], [27]. Its high Km and failure of end-product inhibition make GK a good candidate as a gatekeeper for neuronal glucosensing. Pharmacological inhibition of GK enzymatic activity inhibits GE and stimulates GI neuronal activity at relatively high glucose levels and GK activation simulates GE and inhibits GI activity at low glucose levels [11], [15], [16], [57]. Near complete deletion of GK mRNA in cultured hypothalamic neurons is associated with almost total absence of demonstrable GE and GI neurons which occurs without any appreciable loss of actual viable neurons [15].

Several lines of evidence support a role for hypothalamic GK in the regulation of counterregulatory responses to glucoprivation. Under a variety of conditions in which hypothalamic GK expression (and presumably enzymatic activity) is increased, the neurohumoral counterregulatory responses to insulin-induced hypoglycemia are blunted [11], [43]. The presumed reason for this is that glucosensing neurons, are sensitized to glucose and respond to changes in glucose at much lower levels than they would were GK levels not increased [11], [43]. In addition, the increase of food intake stimulated by insulin-induced hypoglycemia is reduced when hypothalamic GK activity is accentuated by 3rd ventricular injection of a drug which increases GK enzymatic activity. On the other hand, increasing hypothalamic GK activity has no effect on spontaneous feeding in hungry (dark onset) or satiated (light onset) rats (unpublished results). If confirmed, such preliminary data suggest that GK-mediated glucosensing (and possibly glucosensing in general) may be critical for the feeding and the neurohumoral responses to glucoprivation but may not be an important regulator of physiological ingestive behavior.

While changes in blood glucose may not be a critical primary regulator of ingestive behavior, it is likely that glucose levels do play a role in altering the overall activity of metabolic sensing neurons involved in ingestive behavior and energy homeostasis. Glucosensing neurons also express and respond to peripheral hormones such as leptin and insulin which inform the brain as to the metabolic status and adiposity of the body [16], [41], [48], [56]. In such neurons glucose may act as one of several metabolic and hormonal regulators of their membrane potential and final activity. For example, depending upon the ambient glucose levels [49], [56], insulin can hyperpolarize, depolarize or have no effect on their membrane potential. At high, supraphysiological glucose levels, insulin hyperpolarizes the membrane and reduces firing by activating the KATP channel [49]. However, this effect is reduced as ambient glucose levels approach physiological levels and, at glucose levels approximating those seen during hypoglycemia, insulin activates GE neurons by a mechanism that does not involve the KATP channel [56]. This latter effect might reflect an action of insulin on the insulinactivated glucose transporter, GLUT4, since many glucosensing neurons express both insulin receptors and GLUT4 mRNA [16]. Similarly, glucosensing neurons express leptin receptors and the effects of leptin may also be dependent upon ambient glucose concentrations. Some glucosensing neurons also respond to changes in ambient fatty acids levels by altering their activity [25], [55] while others change activity in response to changes in lactate [47], [58] or ketone body availability [31]. Such studies demonstrate the fact that metabolic sensing neurons can respond to a variety of metabolic and hormonal signals from the periphery. On the other hand, no single study has demonstrated that these neurons are specifically involved in the physiologic regulation of energy homeostasis and ingestive behavior.

In conclusion, the brain has evolved specialized groups of neurons which respond to signals from the periphery required for the regulation of ingestive behavior and energy homeostasis. Even 50 years after Jean Mayer postulated that such neurons might act as critical regulators of food intake, we still do not have definitive evidence that glucose acts as a primary regulator of ingestion under physiological conditions. On the other hand, there is good evidence that glucosensing neurons are a key element in the ability of the animal to seek and ingest food during periods of severe glucoprivation. Evolving evidence suggests that the ability of a given neuron to sense and respond to glucose is important in regulating various aspects of energy homeostasis because such neurons also integrate metabolic hormonal and hard-wired neural inputs from the periphery to regulate their activity. In that context, glucose is likely to act as one of many inputs which are summated at the level of the cell's membrane potential. Although not every metabolic sensing neuron responds to all of these signaling molecules, many do appear to respond to one or more such agents. Because signals from the periphery arrive with substantially different temporal profiles, the overall effect on neuronal activity is constantly shifting throughout the day. Thus, such neurons are true metabolic sensors and the combined inputs from these various sources is likely to be the main way in which the brain communicates with the periphery in its role as the primary regulator of energy homeostasis and ingestive behavior. Alterations in the ability of such neurons to sense and respond to signals from the periphery appear to underlie the predisposition of some animals to become obese when fat and caloric density of their diets is increased [8], [21], [22], [23]. Since glucose and fatty acid metabolism are dependent upon central sensing of peripheral signals [35], [36], [37], disordered of metabolic sensing might also underlie central insulin and leptin resistance seen in obesity and type 2 diabetes mellitus. Thus, a better understanding of the mechanisms underlying metabolic sensing might provide us with potential therapeutic modalities for the treatment of these disorders.

References (59)

  • V.H. Routh et al.

    Phosphorylation modulates the activity of the ATP-sensitive K+ channel in the ventromedial hypothalamic nucleus

    Brain Res

    (1997)
  • V.L. Sutherland et al.

    Developmental expression of glucokinase in rat hypothalamus

    Brain Res Dev Brain Res

    (2005)
  • K. Yettefti et al.

    Characteristics of glycemia-sensitive neurons in the nucleus tractus solitarii: Possible involvement in nutritional regulation

    Physiol Behav

    (1997)
  • B.K. Anand et al.

    Hypothalamic control of food intake in rats and cats

    Yale J Biol Med

    (1951)
  • B.K. Anand et al.

    Activity of single neurons in the hypothalamus feeding centers: effect of glucose

    Am J Physiol

    (1964)
  • D. Burdakov et al.

    Physiological changes in glucose differentially modulate the excitability of hypothalamic melanin-concentrating hormone and orexin neurons in situ

    J Neurosci

    (2005)
  • D.J. Clegg et al.

    Reduced anorexic effects of insulin in obesity-prone rats and rats fed a moderate fat diet

    Am J Physiol

    (2005)
  • M. Dallaporta et al.

    Solitary tract nucleus sensitivity to moderate changes in glucose level

    NeuroReport

    (1999)
  • M.G. De Vries et al.

    Extracellular glucose in rat ventromedial hypothalamus during acute and recurrent hypoglycemia

    Diabetes

    (2003)
  • A.A. Dunn-Meynell et al.

    Glucokinase is the likely mediator of glucosensing in both glucose excited and glucose inhibited central neurons

    Diabetes

    (2002)
  • H. Heimberg et al.

    The glucose sensor protein glucokinase is expressed in glucagon-producing alpha-cells

    Proc Natl Acad Sci U S A

    (1996)
  • N. Ibrahim et al.

    Hypothalamic proopiomelanocortin neurons are glucose responsive and express K(ATP) channels

    Endocrinology

    (2003)
  • L. Kang et al.

    Glucokinase is a critical regulator of ventromedial hypothalamic neuronal glucosensing

    Diabetes

    (2006)
  • L. Kang et al.

    Physiological and molecular characteristics of rat hypothalamic ventromedial nucleus glucosensing neurons

    Diabetes

    (2004)
  • D. Kohno et al.

    Ghrelin directly interacts with neuropeptide-Y-containing neurons in the rat arcuate nucleus: Ca2+ signaling via protein kinase A and N-type channel-dependent mechanisms and cross-talk with leptin and orexin

    Diabetes

    (2003)
  • B.E. Levin

    Glucosensing neurons do more than just sense glucose

    Int J Obes Relat Metab Disord

    (2001)
  • B.E. Levin et al.

    Reduced central leptin sensitivity in rats with diet-induced obesity

    Am J Physiol

    (2002)
  • B.E. Levin et al.

    Obesity-prone rats have normal blood–brain barrier transport but defective central leptin signaling prior to obesity onset

    Am J Physiol

    (2003)
  • B.E. Levin et al.

    Abnormalities of leptin and ghrelin regulation in obesity-prone juvenile rats

    Am J Physiol

    (2003)
  • Cited by (106)

    • Dietary influences on cognition

      2018, Physiology and Behavior
    View all citing articles on Scopus
    View full text