Diabetes Volume 65, September 2016 2473 Charles V. Mobbs Orphaned No More? Glucose-Sensing Hypothalamic Neurons Control Insulin Secretion Diabetes 2016;65:2473–2475 | DOI: 10.2337/dbi16-0027 Well before Dr. Minkowski discovered the role of the (4) revised the so-called glucostat hypothesis, which pancreas in regulating blood glucose (1889), Dr. Bernard stated that blood glucose was a key satiety signal that haddiscoveredtheroleofthebrain(specifically the base acted on hypothalamic glucose-sensing neurons and it of the fourth ventricle) in regulating blood glucose was damage to these neurons that mediated obesity (1849). Bernard’s basic observation has been replicated after VMN lesions. A major rationale for this hypothesis by many interventions involving pharmacological manip- was that peripheral injections of gold thioglucose pro- ulations directed toward the brain stem (1). Neverthe- duced obesity and lesions apparently specifictothe less, the overt production of diabetes by the removal of VMN and, interestingly, to the nucleus tractus solitarius the pancreas and the eventual discovery of the determi- (NTS) of the brain stem, lesions completely dependent native role of insulin secreted by pancreatic islets in on the glucose moiety. Motivated in part by these stud- regulating blood glucose by Drs. Banting and Best ies, in 1969 Oomura et al. (5) reported that activity of COMMENTARY (1922) led to an overwhelming emphasis in studying different neurons in the VMN and LHA could be either islet function (and later insulin sensitivity) in the study excited or inhibited by glucose. of blood glucose regulation. Nevertheless, the glucostat hypothesis was gradually At about the same time, the function of the hypothal- marginalized, especially after the Freidman laboratory amus began to be studied extensively because of early discovered leptin and that the signaling form of the leptin evidence that damage to the hypothalamus in humans receptor is largely confined to hypothalamic areas that produced a variety of neuroendocrine disturbances, in- regulate energy balance (6,7). However this marginaliza- cluding obesity (Erdheim 1904). However the relative tion left glucose-sensing neurons orphaned, in search of importanceofthehypothalamusversusthepituitaryin function(s). these systemic pathologies continued to be controversial Addressing this issue, in 1981 Ritter et al. (8) reported until the classic study by Hetherington and Ransom (2) that infusion of the glucose metabolism inhibitor 5-thio- reported that specific lesions in the ventromedial nu- glucose, apparently confined to the caudal brain stem, cleus (VMN) cause robust obesity in rats, an extremely produced hyperglycemia as well as feeding. Furthermore, robust effect replicated in many species, including hu- the Sherwin laboratory reported that the glucose metab- mans. Subsequent extensions of these studies demon- olism inhibitor 2-deoxyglucose, confined to the VMN, also strated that lesions of an area slightly lateral to the VMN produced hyperglycemia and associated counterregulatory (lateral hypothalamic area [LHA]) produced the opposite responses (9), whereas infusion of glucose into the VMN effect: starvation. Furthermore, lesions in the paraven- prevented counterregulatory responses (10). These stud- tricular nucleus (PVN) also produce obesity, although ies have been extensively corroborated, leading to a broad they do not prevent feeding induced by hypoglycemia consensus that glucose-sensing neurons in the VMN and or 2-deoxy-D-glucose (2-DG) (3). brain stem (almost certainly including the NTS) play a key As early as 1916, Carlson proposed that glucose role in glucose regulation. Subsequent studies supported controlled satiety, but his proposed mechanisms in- that in at least some neurons glucose either excites (11) or volving stomach contractions as hunger signals were inhibits (12) glucose-sensing neurons via the putative glucose- eventually largely discarded. However, in 1953 Mayer sensing enzyme, the pancreatic form of glucokinase (pGK). Departments of Neuroscience; Endocrinology, Diabetes and Bone Disease; and © 2016 by the American Diabetes Association. Readers may use this article as Geriatrics and Palliative Medicine, Icahn School of Medicine at Mount Sinai, New long as the work is properly cited, the use is educational and not for profit, and York, NY the work is not altered. More information is available at http://diabetesjournals Corresponding author: Charles V. Mobbs, [email protected]. .org/site/license. See accompanying article, p. 2711. 2474 Commentary Diabetes Volume 65, September 2016 Nevertheless, the mechanisms by which brain glucose- Table 1—Order of connectivity from the brain region sensing neurons regulate glucose homeostasis have indicated to pancreatic islets, from first to last, as indicated remained unclear. This question is robustly addressed by by viral tracing in Rosario et al. (13) the study by Rosario et al. (13) in this issue of Diabetes. Order of Using elegant improved retrograde pseudorabies viral connectivity Metabolic phenotype (PRV) techniques, these investigators traced the connec- 1. NTS/DMX GTG lesion tivity from the brain to the pancreas to b-cells, using an POMC expression inducible mouse insulin promoter Cre-recombinase con- Activated by glucopenia Lesions produce mild hyperphagia struct. Taking advantage of the fact that neurons more directly connected to the b-cells are labeled first, a time 2. PAG Little if any course study convincingly demonstrated the order in 3. PVN Lesion produces hyperphagia and obesity which the neurons connected: the (presumably afferent) Lesion-induced obesity blocked by pair-feeding pathway begins in the hypothalamus (labeled last, the Lesion causes increased blood glucose LHA, arcuate nucleus [ARC], and VMN, through the and insulin PVN), through the periaqueductal gray (PAG) (labeled Activated by hypoglycemia second), through the brain stem (NTS/dorsal motor Low expression of GK nucleus of the vagus (DMX), labeled first), then through 3. LHA Lesions produce starvation and impaired autonomic afferents to b-cells (see Fig. 4C in ref. 13) response to glucopenia (Table 1). Of particular interest, many of the projecting Predominantly glucose inhibited Activated by hypoglycemia neurons from the ARC, VMN, and LHA coexpressed with GK overlaps with projections to pancreas pGK, and a significant subset of these neurons also Enhanced hexokinase decreases glucose, expressed c-fos after intraperitoneal glucose injection. no change in insulin These results convincingly suggest a modulatory role for 4. ARC Neonatal MSG lesions produce modest hypothalamic neurons to regulate b-cell function via this obesity pathway. Adult lesions (AgRP/NPY neurons) by To address the functional significance of this pathway, diphtheria toxin produce starvation AgRP/NPY neurons inhibited by glucose Rosario et al. (13) assessed the effect of overexpressing AgRP/NPY neurons activated by hexokinase I (HK1) in hypothalamic neurons on pancre- hypoglycemia and fasting POMC neurons atic function. A similar manipulation in b-cells basically Activated by glucose, inhibited by fasting mimics the effects of glucose, enhancing basal insulin Enhanced POMC expression reverses diabetes secretion but not glucose-stimulated insulin secretion, al- GK overlaps with projections to pancreas Enhanced hexokinase increases glucose, though overexpression of HK1 enhanced insulin secretion decreases insulin relative to controls even in the stimulated state (14). If 4. VMN Lesion produces hyperphagia, reduced hypothalamic neurons behaved similarly, the expectation metabolism, and obesity would be that overexpression of hypothalamic HK1 would Lesion-induced obesity not prevented by mimic glucose signaling. As overexpression of HK1, at least pair-feeding in the ARC and VMN, reduced insulin secretion rather Lesion produces hyperinsulinemia, not hyperglycemia than enhanced it, the most logical conclusion is that the GTG lesion overlaps with POMC neurons dominant effect of HK1 in the these nuclei was in glucose- Genetic ablation of several VMH genes inhibited neurons, analogous perhaps to glucagon expres- (Sf1, BDNF) produces obesity sion in a-cells. Inhibition of VMN neurons also appears to Lesion blocks counterregulatory response be mediated by pancreatic-like mechanisms, including a to hypoglycemia Local glucopenia produces role for pGK and glycolysis (12). These results are sup- fi counterregulatory responses ported by more speci c activation of pGK-expressing neu- Local infusion of glucose prevents rons in the VMN that also reduced insulin secretion, counterregulatory responses whereas specific inhibition of these neurons stimulated in- GK overlaps with projections to pancreas sulin secretion (15). These glucose-inhibited VMN neurons Enhanced hexokinase decreases insulin, no change in glucose likely mediate counterregulatory responses as they are ac- tivated by 2-DG (12), and lesions of the VMN block coun- Phenotypes as derived by a variety of studies, including the study by Rosario et al. (13). GTG, gold thioglucose; MSG, terregulatory responses to hypoglycemia (16). monosodium glutamate; POMC, proopiomelanocortin. As with all good studies, the study by Rosario et al. (13) should stimulate many further studies. For example, it on pancreatic function. Analyses of other hypothalamic will be of great interest to assess which subset of hypo- neurons would be similarly compelling. thalamic neurons mediates the effects of HK1 on insulin secretion. NPY/AgRP neurons are activated