The mechanisms that underlie glucose sensing during hypoglycaemia in diabetes

Transcription

1 Blackwell Publishing Ltd Review Article DIABETICMedicine DOI: /j x The mechanisms that underlie glucose sensing during hypoglycaemia in diabetes R. McCrimmon Yale University School of Medicine, Department of Internal Medicine, New Haven, CT, USA Accepted 10 October 2007 Abstract Hypoglycaemia is a frequent and greatly feared side-effect of insulin therapy, and a major obstacle to achieving nearnormal glucose control. This review will focus on the more recent developments in our understanding of the mechanisms that underlie the sensing of hypoglycaemia in both non-diabetic and diabetic individuals, and how this mechanism becomes impaired over time. The research focus of my own laboratory and many others is directed by three principal questions. Where does the body sense a falling glucose? How does the body detect a falling glucose? And why does this mechanism fail in Type 1 diabetes? Hypoglycaemia is sensed by specialized neurons found in the brain and periphery, and of these the ventromedial hypothalamus appears to play a major role. Neurons that react to fluctuations in glucose use mechanisms very similar to those that operate in pancreatic B- and A-cells, in particular in their use of glucokinase and the K ATP channel as key steps through which the metabolic signal is translated into altered neuronal firing rates. During hypoglycaemia, glucose-inhibited (GI) neurons may be regulated by the activity of AMP-activated protein kinase. This sensing mechanism is disturbed by recurrent hypoglycaemia, such that counter-regulatory defence responses are triggered at a lower glucose level. Why this should occur is not yet known, but it may involve increased metabolism or fuel delivery to glucose-sensing neurons or alterations in the mechanisms that regulate the stress response. Diabet. Med. 25, (2008) Keywords hypoglycaemia, glucose-excited neurons, glucose-inhibited neurons, ventromedial hypothalamus, AMPactivated protein kinase Abbreviations ACTH, adrenocorticotrophic hormone; AgRP, agouti-related peptide; AMP, adenosine monophosphate; AMPK, AMP-activated protein kinase; APF, action potential frequency; Arc, arcuate nucleus; ATP, adenosine triphosphate; CRH, corticotrophin-releasing hormone; CSF, cerebrospinal fluid; DMH, dorsomedial hypothalamus; ECF, extracellular fluid; GABA, gamma-aminobutyric acid; GE, glucose-excited neuron; GH, growth hormone; GI, glucoseinhibited neuron; GLUT, glucose transporter; GK, glucokinase; HAAF, hypoglycaemia-associated autonomic failure; K ATP, ATP-sensitive potassium channel; KO, knock out; MCT2, monocarboxylate transporter 2; MRNA, messanger ribonucleic acid; NO, nitric oxide; NOS, nitric oxide synthase; POMC, pro-opiomelano cortin; PVN, paraventricular nucleus; rt-pcr, reverse transcription polymerase chain reaction; T1DM, Type 1 diabetes mellitus; VDCC, voltagedependent calcium channel; VMH, ventromedial hypothalamus; VMN, ventromedial nucleus Introduction Shortly after the introduction of insulin in the management of T1DM, clinicians became aware of the potential for insulin therapy to induce iatrogenic hypoglycaemia. Despite the introduction of insulin analogues and improved delivery systems, hypoglycaemia remains the major adverse effect of Correspondence to: Rory J. McCrimmon, Yale University, Department of Internal Medicine Section of Endocrinology, PO Box , New Haven, CT , USA. rory.mccrimmon@yale.edu insulin therapy and has emerged as a major limitation to achieving near-normal glucose control, which is required to reduce the risk of microvascular complications [1]. An increasing awareness of hypoglycaemia in clinical practice provided the initial stimulus to a series of excellent and detailed human physiological studies in the 1980s and 1990s, which greatly increased our understanding of the counter-regulatory defence systems that prevent and correct hypoglycaemia and, importantly, how these differed over time in individuals with T1DM (for a detailed review of this area see [2]). However, in order to understand the mechanisms that underlie glucose sensing, a Journal compilation 2008 Blackwell Publishing Ltd. Diabetic Medicine, 25,

2 DIABETICMedicine Glucose sensing during hypoglycaemia R. McCrimmon FIGURE 1 Neuroendocrine and neuronal pathways involved in generating the counterregulatory response to hypoglycaemia. Sensing of hypoglycaemia by specialized neurons in the hypothalamus leads to: (i) activation of both branches of the autonomic nervous system, culminating (amongst other actions) in the secretion of epinephrine and noradrenaline from the adrenal medulla and in the suppression of insulin release and stimulation of glucagon release from the pancreas, and (ii) stimulation to release both anterior and posterior pituitary hormones such as ACTH and GH. prerequisite to understanding why this homeostatic system becomes defective in T1DM, investigators have in recent years increasingly used more basic in vitro techniques and animal models. This is an area of research that is developing rapidly, and focuses on three basic questions. Where does the body detect hypoglycaemia? How does the body detect hypoglycaemia? And why does this mechanism become defective over time in T1DM? This review will focus on the more recent developments in this field, which are beginning to answer these three important questions. Where does the body sense a falling glucose? Glucose is an integral part of whole-body energy homeostasis and is tightly regulated by numerous endocrine, neuronal and behavioural systems, which ensure that glucose levels in the blood are maintained within a narrow physiological range. It is perhaps not surprising, given the fundamental importance of glucose to the organism, that the capacity to sense glucose is not limited to any single organ system. Apart from the classical glucose sensor, the pancreatic B-cell, glucose-sensing neurons in the periphery have, to date, been found in the intestine [3], hepatoportal vein [4 9] and carotid body [10]. Within the brain, glucose-sensing neurons are found in areas such as the septum [11], amygdala [12], striatum [13], motor cortex [14], hindbrain [15 19] and hypothalamus [20 25] (Fig. 1). It is anticipated, but has not as yet been convincingly shown, that these central and peripheral glucose-sensing neurons form an integrated neural network that monitors and responds to changes in the glucose levels to which they are exposed. The reliance of the brain on glucose to meet its energy demands suggests that, within the context of hypoglycaemia, brain glucose sensing may predominate. Specialized neurons in the hypothalamus and hindbrain [26] are recognized as playing an important role in glucose sensing during hypoglycaemia. Within the hypothalamus, we now recognize that glucosesensing neurons are found in certain distinct regions, namely the VMH [27 30] (which includes the arcuate and ventromedial nuclei), the PVN [31] and the DMH [32]. Perhaps the most studied of these regions is the VMH. Chemical destruction of the VMH in a rodent model with ibotenic acid was shown to cause a ~75% reduction in the hormonal counter-regulatory response to acute hypoglycaemia [29]. Later, it was shown in awake rats that local perfusion of the VMH with 2-deoxyglucose (a non-metabolizable form of glucose that effectively causes local hypoglycaemia ) stimulated a classical systemic counterregulatory response [30], whereas local perfusion of the VMH with glucose during systemic hypoglycaemia markedly suppressed the hormonal counter-regulatory response [28]. Taken together, these three studies provided good evidence for the VMH playing a key role in the sensing of hypoglycaemia. The glucose-sensing neurons share certain common features. Within the brain, they localize to regions adjacent to the third or fourth ventricle or to the circumventricular organs (these are regions of the brain where the blood brain barrier is leaky or absent). This potentially allows glucose-sensing neurons direct sampling, and hence monitoring, of glucose levels in the blood, brain and CSF. This is important because the presence of the blood brain barrier ensures that brain glucose levels are only ~10 30% of the levels seen in the blood [33 35]. Thus, glucose-sensing neurons are able to integrate changes in 514 Journal compilation 2008 Blackwell Publishing Ltd. Diabetic Medicine, 25,

3 Review article DIABETICMedicine glucose within each of these different regions. In addition, glucose-sensing neurons are generally located in areas involved in the control of the autonomic nervous system, neuroendocrine function, nutrient metabolism and energy homeostasis. What this effectively means is that brain glucose-sensing neurons are able to receive inputs from blood nutrient levels and a variety of endocrine hormones, as well as from numerous peripheral and central sensory systems. This information can then be synthesized and used to generate the appropriate physiological response, the ultimate aim of which is to preserve glucose homeostasis. How does the body detect a falling glucose? In 1953, Jean Mayer proposed the glucostatic hypothesis [36]. Hypothalamic glucoreceptors, it was proposed, could sense fluctuations in glucose and translate that signal into a change in neural activity [36]. The defining feature of these neurons should be their use of glucose, not simply as a fuel, but as a signalling molecule that regulates their activity. All neurons use glucose as their major fuel source and all neurons will ultimately alter their firing rates when glucose homeostasis is significantly disrupted, but glucose-sensing neurons alter their membrane potential, action potential frequency and/or rate of neurotransmitter release over the more physiological ranges of glucose to which they are exposed. Such specialized neurons were first demonstrated by Oomura and colleagues in 1969 [22]. These neurons are glucose -sensing in so far as glucose is the major metabolic substrate for the brain, but the fact that these neurons can use other fuels, such as lactate produced either by astrocytes [37 39], delivered locally [40,41] or systemically [42 45] to alter their function, suggests that it is more likely to be intracellular ATP that determines the activity of these neurons. This is intriguing because neuronal levels of ATP are generally thought to be well maintained, thus it is also likely that subcellular compartmentalization of the glucosesensing apparatus must exist to provide sensing capability. Such compartmentalization has been shown within pancreatic β-cells [46]. GE neurons There are thought to be two predominant subtypes of glucosesensing neurons, namely GE neurons (whose activity increases as glucose levels rise) and GI neurons (whose activity decreases as glucose levels rise) [47]. The mechanisms through which these neurons sense alterations in glucose remain incompletely understood. Classical glucose sensing in the pancreatic islet served as a model for the initial research in this field [48]. In the pancreatic β-cell, glucose is transported into the cell through a high-capacity glucose transporter (GLUT) to allow rapid equilibration of extracellular and cytosolic glucose. The rate of glucose metabolism, which is closely coupled to insulin secretion, is determined by glucokinase (GK), the enzyme in the rate-limiting step of glycolysis [49]. Metabolism of glucose leads to several intracellular events, the culmination of which is an increase in cytosolic calcium. One of the key steps in this signalling pathway is the depolarization of the B-cell that follows closure of ATP-sensitive potassium (K ATP ) channels in the plasma membrane [48]. The exciting discovery of glucokinase and K ATP channels in glucose-sensing regions in the brain lead to the hypothesis that these enzymes might also serve key roles in glucose sensing, particularly in GE neurons [48,50]. K ATP channels have been demonstrated throughout the brain, including hypothalamic regions thought to be involved in glucose sensing [27,51 53]. Using single-cell rt-pcr to analyse glucose-sensing neurons (identified electrophysiologically in a hypothalamic slice preparation), investigators have shown that they express mrna for Kir6.2 and SUR1, the two subunits that comprise the K ATP channel in the pancreatic B-cell [54]. In addition, electrophysiological studies of the rat [27,55 57] and mouse [58] have demonstrated that sulphonylureas (agents that block the K ATP channel) can alter the response of GE neurons to changes in ambient glucose, and Kir6.2 KO mice show impaired glucose counter-regulation to systemic hypoglycaemia [58]. Finally, in vivo perfusion of the VMH of rodents with glibenclamide (a K ATP channel blocker) suppresses [59], whereas diazoxide (a K ATP channel opener) amplifies [60], hormonal counter-regulatory responses to acute hypoglycaemia. However, K ATP channels are present on many different neurons and so while they may be required for glucose sensing they are unlikely to have a regulatory role. In the pancreatic B-cell, GK is the critical regulator of glycolytic production of ATP and K ATP channel activity [49]. The pancreatic form of GK is also present in areas of the brain involved in glucose sensing [61 63]. GK mrna is expressed in ~70% of GE neurons [54], and glucokinase has also been shown to play a regulatory role in GE neuron-sensing ability [54,57,61,64]. In addition, the selective down-regulation of GK in GE neurons in primary VMH neuronal cultures led to the loss of all demonstrable GE and GI neurons [65]. However, extracellular levels of glucose are only about 10 30% of the levels found in blood. Microdialysis studies in rats [33,34,66] and human subjects [67] have shown that the ECF glucose levels to which neurons are exposed are in the range of 1 2 mm, and fall to a similar degree during acute hypoglycaemia (~0.5 mm) [33]. Both euglycaemic and hypoglycaemia ECF glucose levels are beneath the range of glucose levels in which GK usually acts in a regulatory manner, thus for GK to perform this role in glucose-sensing neurons it almost certainly has to operate within a distinct subcellular compartment. Glucose signalling in the pancreatic B-cell also requires the GLUT Type 2 (GLUT-2). GLUT-2 mrna has been found in brain glucose-sensing regions [54,68]; in transgenic mice, central GLUT-2 has been shown to be involved in the counterregulatory response to hypoglycaemia. Intriguingly, reintroducing GLUT-2 to glial cells, but not to neurons, restored the defective counter-regulatory response to hypoglycaemia [69]. Taken together, these studies provide good evidence that the glucose-sensing mechanism used by the GE neuron has many similarities with the pancreatic β-cell. In particular, GK and Journal compilation 2008 Blackwell Publishing Ltd. Diabetic Medicine, 25,

4 DIABETICMedicine Glucose sensing during hypoglycaemia R. McCrimmon FIGURE 2 Hypothetical sensing mechanism of GE neurons. Glucose enters the GE neuron through GLUT 2 or 3, and is phosphorylated by GK, acting as the gatekeeper, and regulating the production of cytosolic ATP in a subcellular compartment. The ATP closes K ATP channels in the plasma membrane, causing depolarization. In turn, this leads to Ca 2+ influx through VDCCs, stimulating neurotransmitter release and/or increased APF. Lactate, produced locally by astrocytes or arriving systemically, enters the neuron via MCT2 and can then be metabolized to form ATP. FIGURE 3 Hypothetical sensing mechanism of GI neurons. In GI neurons, GK may once again act as the gatekeeper. A falling glucose results in an increase in the AMP : ATP ratio, an effect that can be mimicked pharmacologically by AICAR. AMPK, once activated, stimulates the formation of NOS, which may diffuse out to adjacent neurons or glial cells. In addition, AMPK may act on chloride channels leading to neuronal depolarization and neurotransmitter release and/or increased APF. This action of AMPK can be blocked pharmacologically by compound C. the K ATP channel appear to be key points through which an increase in the ambient glucose leads to altered firing of GE neurons (Fig. 2). GI neurons In contrast, GI neurons show a decrease in activity as glucose levels rise [47]. It is perhaps easier to think of GI neurons as those glucose-sensing neurons that become more active when glucose levels fall, and as such they may use signalling mechanisms more relevant to the pancreatic A-cell. Unfortunately, like the A-cell, the signalling pathways used by the GI neuron are not well understood (Fig. 3). This may reflect the fact that they are few in number, comprising only 3 14% of neurons in the ventromedial hypothalamic nucleus [61,70]. However, recent evidence suggests that GI neurons may be more prevalent: when improved slice techniques are used in conjunction with novel methods for identifying GI neurons (e.g. membrane potential dyes), as many as ~30 40% of all neurons in the VMH are reported to be GI neurons [71]. GK mrna is expressed in around 40% of GI neurons, and may also serve a regulatory role in these neurons. [65]. More recently, evidence has also emerged that AMP-activated protein kinase (AMPK) plays a key role in the sensing pathway used by GI neurons (Fig. 3). AMPK has been described as an intracellular fuel gauge in that it is activated in response to a rise in the intracellular ratio of AMP to ATP and acts to switch off energy-consuming anabolic processes and switch on energyproducing catabolic processes [72]. Within the brain, AMPK expression is thought to be predominantly neuronal in distribution, with very little expression evident in astrocytes [73]. Canabal et al. [71] demonstrated that in VMN GI neurons exposed to 2.5 mm glucose, AICAR (an activator of AMPK), mimicked the excitatory effect of low glucose (0.5 mm) on action potential frequency. Both low glucose and AICAR were shown to mediate their effects, in part, through an increase in nitric oxide (NO) production in GI neurons. Conversely, increased NO production in response to a low glucose was blocked by compound C, an inhibitor of AMPK [71]. In a related series of studies, Mountjoy et al. [74] reported that activation of AMPK with AICAR or inhibition with compound C altered neuronal activity in GI, but not GE, neurons in an ex vivo hypothalamic cell culture system obtained from the mediobasal (including VMN and Arc) hypothalamus. In this study, AMPK had no effect on the K ATP channel [74], but others have suggested that AMPK may instead act on a chloride channel to depolarize the plasma membrane [75]. In rodent studies, in vivo pharmacological activation of AMPK in the VMH during acute hypoglycaemia amplifies the glucose counter-regulatory response in normal Sprague-Dawley rats [76], and restores the hormonal counter-regulatory response to hypoglycaemia in rats with defective counter-regulation [77]. Conversely, down-regulation of AMPK in the VMH using specific RNA interference suppresses the counter-regulatory response to acute hypoglycaemia [78]. In addition, mice that selectively lack the alpha-catalytic subunit of AMPK in POMC or AgRP neurons (both neuronal types are located in the medio-basal hypothalamus and are involved in feeding 516 Journal compilation 2008 Blackwell Publishing Ltd. Diabetic Medicine, 25,

5 Review article DIABETICMedicine behaviour and energy homeostasis) lose the responsiveness of these neurons to changes in extracellular glucose [79]. What these studies all indicate is that the brain contains specialized neurons that exist in distinct locations and are able to monitor and react to alterations in the glucose concentration to which they are exposed. These neurons, by virtue of specific sensing systems, translate the rate or quantity of glucose oxidation into a neural signal that alters neuronal firing rates and, in the case of hypoglycaemia, leads to the stimulation of a systemic glucose counter-regulatory defence response. The glucose-sensing neurons seem to use signalling mechanisms that are very similar to those used by pancreatic B- and A-cells. The GE neuron, which seems to operate more under conditions of eu- or hyperglycaemia, may use glucokinase as its key regulatory step, and the K ATP channel to then translate that signal into altered neuronal firing rates. In contrast, while the GI neuron (which operates more under hypoglycaemic conditions) may still use glucokinase as a regulatory enzyme, it appears more dependent on alterations in intracellular AMPK activity that, in turn, may act via a chloride channel to alter neuronal firing rates. It is highly likely that these two neuronal populations communicate with each other, perhaps via the inhibitory neurotransmitter GABA [80], to co-ordinate their responses to a changing blood glucose. Potentially, the counterbalance between GI and GE neuronal activity forms the most sensitive means of regulating and maintaining blood glucose within a narrow physiological range and ensuring an adequate supply of glucose to the brain. Counter-regulation in T1DM In health, a fall in plasma glucose level is rapidly detected and a sequence of counter-regulatory responses triggered, which mainly involve: (1) suppression of insulin secretion; (2) counterregulatory hormone release, which rapidly promotes endogenous glucose production and limits peripheral glucose utilization; and (3) subjective awareness of hypoglycaemia. In T1DM, these compensatory systems are disrupted at every level. Firstly, for most individuals, insulin delivery from its subcutaneous depot is unregulated and continues despite the development of hypoglycaemia. Secondly, most individuals with T1DM will over time develop defects in the hormonal counter-regulatory defence against hypoglycaemia. Within 5 years of diagnosis of T1DM, hypoglycaemia fails to stimulate the release of glucagon, the major counter-regulatory hormone (Table 1) [81]. As a result, individuals with T1DM are particularly dependent on the sympathoadrenal (primarily adrenaline and noradrenaline) response to low blood glucose. However, within 10 years of diagnosis, the majority of patients develop additional impairments in sympathoadrenal and other neurohormonal responses against hypoglycaemia (Table 1) [81]. In addition, symptom awareness becomes impaired in individuals with T1DM. The term hypoglycaemiaassociated autonomic failure (HAAF) was introduced by Cryer to incorporate both defective glucose counter-regulation Table 1 Percentage of individuals with T1DM showing deficient responses in each of the major hypoglycaemic counter-regulatory hormones over time. From Mokan et al. [81] Duration of diabetes Glucagon (%) Epinephrine (%) Cortisol (%) GH (%) < 1 year years years > 10 years and hypoglycaemia unawareness [1]. The presence of HAAF greatly increases the risk of suffering severe hypoglycaemia in patients with T1DM [82]. The major factor leading to the loss of glucagon secretion in response to acute hypoglycaemia in T1DM appears to be the loss of insulin secretion from the B-cell. This is because the glucagon-secretory defect is selective (responses to other stimuli such as arginine remain intact), and develops concomitantly with progressive B-cell loss in T1DM [83]. Based on earlier rat anatomical studies suggesting that blood flow in the islet was predominantly from B- to A-cell, it was proposed that insulin under ordinary conditions tonically inhibited glucagon secretion by the pancreatic A-cell [84 86]. Hence, the switch-off hypothesis posited that during hypoglycaemia when insulin secretion is suppressed, glucagon is released from the A-cell. In T1DM, pancreatic B-cells are destroyed in the autoimmune inflammatory process such that this switch-off of insulin no longer occurs. In fact, intra-islet insulin levels may actually rise, leading to the suppression of glucagon secretion [84 86]. There is good evidence in support of this hypothesis, from both in vitro [85 89] and in vivo [89] rodent studies. However, the pancreatic α-cell has an extensive and complex autonomic innervation (reviewed in detail in Taborsky et al. [90]) with considerable redundancy, so while insulin may be the primary factor responsible for this defect, an additional role for alterations in the autonomic nervous system cannot be excluded. The almost universal defect in glucagon secretion means that individuals with T1DM are very reliant on the sympathoadrenal response to hypoglycaemia to restore blood-glucose levels. Unfortunately, this response also appears to become defective over time in T1DM [81]. The major reason for the development of this defect is thought to be the experience of hypoglycaemia itself. In the Diabetes Control and Complications Trial, which compared intensive with standard insulin therapy in T1DM, the major single factor predicting severe hypoglycaemia in the study cohort was reporting a previous episode of severe hypoglycaemia [91]. Intensive insulin therapy, which is associated with more frequent hypoglycaemia, was shown to lower the glucose level at which hormonal counter-regulation was initiated in T1DM [92]. Hypoglycaemia per se is the likely cause of this and it has been shown in both humans [93 95] and rodents [77,96 97] that single or multiple episodes of Journal compilation 2008 Blackwell Publishing Ltd. Diabetic Medicine, 25,

6 DIABETICMedicine Glucose sensing during hypoglycaemia R. McCrimmon antecedent hypoglycaemia lead to suppressed adrenaline responses to a subsequent episode of hypoglycaemia. Data gleaned from studies in rodents suggest that this defect arises through changes in key brain glucose-sensing regions such as the VMH. Recurrent hypoglycaemia has been shown to markedly suppress the counter-regulatory response to 2- deoxyglucose (a non-metabolized form of glucose) perfused into the VMH of rats [98]. It has also been shown that the glucose level at which you first see activation of glucoseinhibited neurons in the VMH is lower in rats that have experienced prior recurrent hypoglycaemia [99]. Moreover, in vivo activation of AMPK in the VMH of rats who have experienced prior recurrent hypoglycaemia can restore hormonal counterregulatory responses to a subsequent hypoglycaemic challenge [77]. Similarly, opening of K ATP channels in the VMH of recurrently hypoglycaemic rats leads to the restoration of normal counter-regulatory responses to subsequent hypoglycaemia. If our model is correct, these studies suggest that the balance between GE and GI neuronal activity is altered by recurrent hypoglycaemia. We would anticipate that GE neurons (acting to suppress counter-regulation) are more likely to be active and/or that GI neurons (acting to amplify counter-regulation) are less likely to be active following recurrent hypoglycaemia. The net effect is that full glucose counter-regulation is initiated at a lower glucose level. This gives individuals less time to react to, and seek treatment for, hypoglycaemia. Why does this mechanism fail in Type 1 diabetes? The question we then need to ask is: why does this happen? The most obvious explanation is that following recurrent hypoglycaemia, glucose-sensing neurons are seeing higher glucose and/or ATP levels during a subsequent hypoglycaemic challenge. This change may arise through an increased supply of fuel through altered glucose [100], or alternate fuel transport [101]. However, while these changes suggest an increased capacity to transport fuels across the blood brain barrier and into glucose-sensing neurons, the data from human studies have been mixed, with some indicating increased whole-brain glucose [102], or acetate [103], uptake after chronic hypoglycaemia and others showing no change [104]. Another possibility is that the brain is able to obtain additional metabolic substrates from more local sources. Recently, it has emerged that brain glycogen may act as a fuel reserve during acute hypoglycaemia [105]. Moreover, brain glycogen levels may actually increase in response to an acute episode of hypoglycaemia [105,106]. This super-compensation of brain glycogen may then provide the additional fuel reserve that leads to defective glucose sensing during a subsequent episode of hypoglycaemia. This theory is attractive, but against it is the fact that brain glycogen levels appear to be very low and a fraction of the levels seen in other tissues such as muscle and liver. Also, super-compensation of glycogen may actually have arisen because of the model employed in these studies where pre-study infusions of glucose and insulin, rather than hypoglycaemia per se, may have led to the described phenomenon. Further research is needed in this area. It has also emerged that mechanisms for regulating the magnitude of the neuroendocrine response to acute hypoglycaemia exist within the brain. The CRH family of ligands and receptors form an ancient and highly conserved means of regulating the neuroendocrine stress response. Within the brain, at key sites involved with autonomic activation, CRH acting through the CRH-R1 receptor appears to lead to an amplification of the autonomic response to stress whereas Urocortin (part of the CRH family of peptides), acting through the CRH-R2 receptor, suppresses the autonomic response to stress. We were recently able to show that such a mechanism operated in the VMH, with urocortin locally delivered to the VMH markedly suppressing the counter-regulatory response to acute hypoglycaemia [107], whereas CRH amplified the response [108]. Moreover, unpublished work from our laboratory would suggest that this counterbalanced system might change following recurrent hypoglycaemia, with an upregulation of CRH-R2-mediated suppressive effects in the VMH. The net effect would be to suppress the counterregulatory response to subsequent hypoglycaemia. We should not be surprised that there is no single answer to the phenomenon of HAAF. Hypoglycaemia represents a profound physiological stress and is likely to activate numerous physiological defence systems that affect both biological and behavioural responses. Our understanding of these physiological defences remains limited, but it is reasonable to conclude that recurrent hypoglycaemia may lead to an increase in the capacity of glucose-sensing regions of the brain to use glucose and/or alternate fuels, as well as to a change in the mechanisms that fine-tune the hypoglycaemic stress response. The net effect of these changes is to suppress the glucose counter-regulatory response to a subsequent episode of hypoglycaemia. What relevance are studies in rodents to human disease? The ultimate aim of all basic research is its translation into the human model. In hypoglycaemia research, electrophysiological studies of hypothalamic slice preparations and rodent models offer much, but one must always be wary of over-interpreting data. The limitations of cell-based and rodent studies are well recognized. However, this does not make the work irrelevant to human disease. Glucose homeostasis, in particular, is fundamental to the survival of a vast range of species from single-cell organisms to humans. AMPK, for instance, is a critical part of the process whereby the yeast cell responds to a period of fuel deprivation, and likewise in rodents clearly plays a major role in glucose sensing during hypoglycaemia. Thus, it appears likely that such ancient and conserved systems may be important in human disease. Another example is glucokinase, which, as reviewed earlier, has been shown in rodents to play a regulatory role in glucose sensing in hypothalamic neurons. 518 Journal compilation 2008 Blackwell Publishing Ltd. Diabetic Medicine, 25,

7 Review article DIABETICMedicine This finding led to human studies where fructose, a modulator of glucokinase activity, was delivered systemically during hypoglycaemia and was shown to amplify the counter-regulatory response in both non-diabetic [109] and T1DM [110] subjects. Finally, the recognition that recurrent hypoglycaemia might lead to adaptive changes that increase the ability of the brain to use alternate fuels has prompted trials in T1DM subjects using medium-chain fatty acids that may provide neuroprotection, particularly during the night, while not leading to a deterioration in overall glucose control. In summary, our understanding of where and how glucose is sensed has increased significantly in the last 20 years. We now recognize that a network of specialized neurons exist that continually monitor the ambient glucose level and that these neurons act in such a way as to try to maintain glucose homeostasis. These neurons appear to use sensing mechanisms very similar to those used within the pancreatic islet, with GK, the K ATP channel and AMPK all playing key roles in central glucose sensing. While this information has come largely from rodent studies using a combination of in vitro and in vivo techniques, it is anticipated that newer imaging techniques such as positron emission tomography and magnetic resonance spectroscopy will allow the translation of these findings into the human model and determine their relevance to human disease. The development of novel therapies or treatment strategies designed to minimize the risk of hypoglycaemia in insulin-treated individuals with diabetes relies on the information that these studies will provide. Competing interests None to declare. Acknowledgements The author acknowledges the post-doctoral fellows and technical staff who contributed greatly to the research that underpins this review. The research work of the author is supported by a research grant from NIDDK (69831), and a Career Development Award from the Juvenile Diabetes Research Foundation. References 1 Cryer PE. Banting Lecture. Hypoglycemia: the limiting factor in the management of IDDM. Diabetes 1994; 43: Frier BM, Fisher BM. Hypoglycaemia and Diabetes: Clinical and Physiological Aspects. London: Edward Arnold, Thorens B, Larsen PJ. Gut-derived signaling molecules and vagal afferents in the control of glucose and energy homeostasis. Curr Opin Clin Nutr Metab Care 2004; 7: Burcelin R, Crivelli V, Perrin C, Da Costa A, Mu J, Kahn BB et al. GLUT4, AMP kinase, but not the insulin receptor, are required for hepatoportal glucose sensor-stimulated muscle glucose utilization. J Clin Invest 2003; 111: Burcelin R, Da Costa A, Drucker D, Thorens B. Glucose competence of the hepatoportal vein sensor requires the presence of an activated glucagon-like peptide-1 receptor. Diabetes 2001; 50: Burcelin R, Dolci W, Thorens B. Glucose sensing by the hepatoportal sensor is GLUT2-dependent: in vivo analysis in GLUT2-null mice. Diabetes 2000; 49: Donovan CM. Portal vein glucose sensing. Diabetes Nutr Metab 2002; 15: Donovan CM, Hamilton-Wessler M, Halter JB, Bergman RN. Primacy of liver glucosensors in the sympathetic response to progressive hypoglycemia. Proc Natl Acad Sci USA 1994; 91: Hevener AL, Bergman RN, Donovan CM. Novel glucosensor for hypoglycemic detection localized to the portal vein. Diabetes 1997; 46: Pardal R, Lopez-Barneo J. Low glucose-sensing cells in the carotid body. Nat Neurosci 2002; 5: Shoji S. Glucose regulation of synaptic transmission in the dorsolateral septal nucleus of the rat. Synapse 1992; 12: Nakano Y, Oomura Y, Lenard L, Nishino H, Aou S, Yamamoto T et al. Feeding-related activity of glucose-and morphine-sensitive neurons in the monkey amygdala. Brain Res 1986; 399: Lee K, Dixon AK, Freeman TC, Richardson PJ. Identification of an ATP-sensitive potassium channel current in rat striatal cholinergic interneurones. J Physiol 1998; 510: Lee K, Dixon AK, Rowe IC, Ashford ML, Richardson PJ. The highaffinity sulphonylurea receptor regulates KATP channels in nerve terminals of the rat motor cortex. J Neurochem 1996; 66: Adachi A, Kobashi M, Miyoshi N, Tsukamoto G. Chemosensitive neurons in the area postrema of the rat and their possible functions. Brain Res Bull 1991; 26: Amoroso S, Schmid-Antomarchi H, Fosset M, Lazdunski M. Glucose, sulfonylureas, and neurotransmitter release: role of ATPsensitive K+ channels. Science 1990; 247: During MJ, Leone P, Davis KE, Kerr D, Sherwin RS. Glucose modulates rat substantia nigra GABA release in vivo via ATP-sensitive potassium channels. J Clin Invest 1995; 95: Finta EP, Harms L, Sevcik J, Fischer HD, Illes P. Effects of potassium channel openers and their antagonists on rat locus coeruleus neurones. Br J Pharmacol 1993; 109: Karschin A, Brockhaus J, Ballanyi K. KATP channel formation by the sulphonylurea receptors SUR1 with Kir6.2 subunits in rat dorsal vagal neurons in situ. J Physiol 1998; 509: Funahashi H, Yada T, Muroya S, Takigawa M, Ryushi T, Horie S et al. The effect of leptin on feeding-regulating neurons in the rat hypothalamus. Neurosci Lett 1999; 264: Nagai K, Niijima A, Nagai N, Hibino H, Chun SJ, Shimizu K et al. Bilateral lesions of the hypothalamic suprachiasmatic nucleus eliminated sympathetic response to intracranial injection of 2- deoxy-d-glucose and VIP rescued this response. Brain Res Bull 1996; 39: Oomura Y, Ono T, Ooyama H, Wayner MJ. Glucose and osmosensitive neurones of the rat hypothalamus. Nature 1969; 222: Orsini JC, Himmi T, Wiser AK, Perrin J. Local versus indirect action of glucose on the lateral hypothalamic neurons sensitive to glycemic level. Brain Res Bull 1990; 25: Orsini JC, Wiser AK, Himmi T, Boyer A, Perrin J. Sensitivity of lateral hypothalamic neurons to glycemic level: possible involvement of an indirect adrenergic mechanism. Brain Res Bull 1991; 26: Silver IA, Ereciska M. Glucose-induced intracellular ion changes in sugar-sensitive hypothalamic neurons. J Neurophysiol 1998; 79: Ritter S, Dinh TT, Zhang Y. Localization of hindbrain glucoreceptive sites controlling food intake and blood glucose. Brain Res 2000; 856: Journal compilation 2008 Blackwell Publishing Ltd. Diabetic Medicine, 25,

8 DIABETICMedicine Glucose sensing during hypoglycaemia R. McCrimmon 27 Ashford ML, Boden PR, Treherne JM. Tolbutamide excites rat glucoreceptive ventromedial hypothalamic neurones by indirect inhibition of ATP-K+ channels. Br J Pharmacol 1990; 101: Borg MA, Sherwin RS, Borg WP, Tamborlane WV, Shulman GI. Local ventromedial hypothalamus glucose perfusion blocks counterregulation during systemic hypoglycemia in awake rats. J Clin Invest 1997; 99: Borg WP, During MJ, Sherwin RS, Borg MA, Brines ML, Shulman GI. Ventromedial hypothalamic lesions in rats suppress counterregulatory responses to hypoglycemia. J Clin Invest 1994; 93: Borg WP, Sherwin RS, During MJ, Borg MA, Shulman GI. Local ventromedial hypothalamus glucopenia triggers counterregulatory hormone release. Diabetes 1995; 44: Evans SB, Wilkinson CW, Gronbeck P, Bennett JL, Taborsky GJ Jr, Figlewicz DP. Inactivation of the PVN during hypoglycemia partially simulates hypoglycemia-associated autonomic failure. Am J Physiol- Regul Integr Comp Physiol 2003; 284: R Evans SB, Wilkinson CW, Gronbeck P, Bennett JL, Zavosh A, Taborsky GJ Jr et al. Inactivation of the DMH selectively inhibits the ACTH and corticosterone responses to hypoglycemia. Am J Physiol-Regul Integr Comp Physiol 2004; 286: R de Vries MG, Arseneau LM, Lawson ME, Beverly JL. Extracellular glucose in rat ventromedial hypothalamus during acute and recurrent hypoglycemia. Diabetes 2003; 52: McNay EC, Gold PE. Extracellular glucose concentrations in the rat hippocampus measured by zero-net-flux: effects of microdialysis flow rate, strain, and age. J Neurochem 1999; 72: Silver IA, Erecinska M. Extracellular glucose concentration in mammalian brain: continuous monitoring of changes during increased neuronal activity and upon limitation in oxygen supply in normo-, hypo-, and hyperglycemic animals. J Neurosci 1994; 14: Mayer J. Glucostatic mechanism of regulation of food intake. New Engl J Med 1953; 249: Magistretti PJ, Sorg O, Naichen Y, Pellerin L, de Rham S, Martin JL. Regulation of astrocyte energy metabolism by neurotransmitters. Renal Physiol Biochem 1994; 17: Magistretti PJ, Sorg O, Yu N, Martin JL, Pellerin L. Neurotransmitters regulate energy metabolism in astrocytes: implications for the metabolic trafficking between neural cells. Dev Neurosci 1993; 15: Pellerin L, Magistretti PJ. Glutamate uptake into astrocytes stimulates aerobic glycolysis: a mechanism coupling neuronal activity to glucose utilization. Proc Natl Acad Sci USA 1994; 91: Song Z, Routh VH. Differential effects of glucose and lactate on glucosensing neurons in the ventromedial hypothalamic nucleus. Diabetes 2005; 54: Borg MA, Tamborlane WV, Shulman GI, Sherwin RS. Local lactate perfusion of the ventromedial hypothalamus suppresses hypoglycemic counterregulation. Diabetes 2003; 52: King P, Kong MF, Parkin H, MacDonald IA, Barber C, Tattersall RB. Intravenous lactate prevents cerebral dysfunction during hypoglycaemia in insulin-dependent diabetes mellitus. Clin Sci 1998; 94: Maran A, Cranston I, Lomas J, Macdonald I, Amiel SA. Protection by lactate of cerebral function during hypoglycaemia. Lancet 1994; 343: Thurston JH, Hauhart RE. Effect of momentary stress on brain energy metabolism in weanling mice: apparent use of lactate as cerebral metabolic fuel concomitant with a decrease in brain glucose utilization. Metab Brain Dis 1989; 4: Thurston JH, Hauhart RE, Schiro JA. Lactate reverses insulininduced hypoglycemic stupor in suckling-weanling mice: biochemical correlates in blood, liver, and brain. J Cereb Blood Flow Metab 1983; 3: Kennedy HJ, Pouli AE, Ainscow EK, Jouaville LS, Rizzuto R, Rutter GA. Glucose generates sub-plasma membrane ATP microdomains in single islet beta-cells. Potential role for strategically located mitochondria. J Biol Chem 1999; 274: Routh VH. Glucose-sensing neurons: are they physiologically relevant? Physiol Behav 2002; 76: Schuit FC, Huypens P, Heimberg H, Pipeleers DG. Glucose sensing in pancreatic beta-cells: a model for the study of other glucoseregulated cells in gut, pancreas, and hypothalamus. Diabetes 2001; 50: Matschinsky FM. Banting Lecture A lesson in metabolic regulation inspired by the glucokinase glucose sensor paradigm. Diabetes 1996; 45: Levin BE, Routh VH, Kang L, Sanders NM, Dunn-Meynell AA. Neuronal glucosensing: what do we know after 50 years? Diabetes 2004; 53: Ashford ML, Boden PR, Treherne JM. Glucose-induced excitation of hypothalamic neurones is mediated by ATP-sensitive K+ channels. Pflugers Archiv-Eur J Physiol 1990; 415: Dunn-Meynell AA, Rawson NE, Levin BE. Distribution and phenotype of neurons containing the ATP-sensitive K+ channel in rat brain. Brain Res 1998; 814: Dunn-Meynell AA, Routh VH, McArdle JJ, Levin BE. Low-affinity sulfonylurea binding sites reside on neuronal cell bodies in the brain. Brain Res 1997; 745: Kang L, Routh VH, Kuzhikandathil EV, Gaspers LD, Levin BE. Physiological and molecular characteristics of rat hypothalamic ventromedial nucleus glucosensing neurons. Diabetes 2004; 53: Dallaporta M, Perrin J, Orsini JC. Involvement of adenosine triphosphate-sensitive K+ channels in glucose-sensing in the rat solitary tract nucleus. Neurosci Lett 2000; 278: Spanswick D, Smith MA, Groppi VE, Logan SD, Ashford ML. Leptin inhibits hypothalamic neurons by activation of ATP-sensitive potassium channels. Nature 1997; 390: Yang XJ, Kow LM, Funabashi T, Mobbs CV. Hypothalamic glucose sensor. Similarities to and differences from pancreatic beta-cell mechanisms. Diabetes 1999; 48: Miki T, Liss B, Minami K, Shiuchi T, Saraya A, Kashima Y et al. ATP-sensitive K+ channels in the hypothalamus are essential for the maintenance of glucose homeostasis. Nat Neurosci 2001; 4: Evans ML, McCrimmon RJ, Flanagan DE, Keshavarz T, Fan X, McNay EC et al. Hypothalamic ATP-sensitive K+ channels play a key role in sensing hypoglycemia and triggering counterregulatory epinephrine and glucagon responses. Diabetes 2004; 53: McCrimmon RJ, Evans ML, Fan X, McNay EC, Chan O, Ding Y et al. Activation of KATP channels in the ventromedial hypothalamus amplifies counterregulatory hormone responses to hypoglycemia in normal and recurrently hypoglycemic rats. Diabetes 2005; 54: Dunn-Meynell AA, Routh VH, Kang L, Gaspers L, Levin BE. Glucokinase is the likely mediator of glucosensing in both glucoseexcited and glucose-inhibited central neurons. Diabetes 2002; 51: Jetton TL, Liang Y, Pettepher CC, Zimmerman EC, Cox FG, Horvath K et al. Analysis of upstream glucokinase promoter activity in transgenic mice and identification of glucokinase in rare neuroendocrine cells in the brain and gut. J Biol Chem 1994; 269: Lynch RM, Tompkins LS, Brooks HL, Dunn-Meynell AA, Levin BE. Localization of glucokinase gene expression in the rat brain. Diabetes 2000; 49: Wang R, Liu X, Hentges ST, Dunn-Meynell AA, Levin BE, Wang W et al. The regulation of glucose-excited neurons in the hypothalamic 520 Journal compilation 2008 Blackwell Publishing Ltd. Diabetic Medicine, 25,

9 Review article DIABETICMedicine arcuate nucleus by glucose and feeding-relevant peptides. Diabetes 2004; 53: Kang L, Dunn-Meynell AA, Routh VH, Gaspers LD, Nagata Y, Nishimura T et al. Glucokinase is a critical regulator of ventromedial hypothalamic neuronal glucosensing. Diabetes 2006; 55: McNay EC, Gold PE. Age-related differences in hippocampal extracellular fluid glucose concentration during behavioral testing and following systemic glucose administration. J Gerontol Series A-Biol Sci Med Sci 2001; 56: B Abi-Saab WM, Maggs DG, Jones T, Jacob R, Srihari V, Thompson J et al. Striking differences in glucose and lactate levels between brain extracellular fluid and plasma in conscious human subjects: effects of hyperglycemia and hypoglycemia. J Cereb Blood Flow Metab 2002; 22: Leloup C, Arluison M, Lepetit N, Cartier N, Marfaing-Jallat P, Ferre P et al. Glucose transporter 2 (GLUT 2): expression in specific brain nuclei. Brain Res 1994; 638: Marty N, Dallaporta M, Foretz M, Emery M, Tarussio D, Bady I et al. Regulation of glucagon secretion by glucose transporter Type 2 (glut2) and astrocyte-dependent glucose sensors. J Clin Invest 2005; 115: Song Z, Levin BE, McArdle JJ, Bakhos N, Routh VH. Convergence of pre- and postsynaptic influences on glucosensing neurons in the ventromedial hypothalamic nucleus. Diabetes 2001; 50: Canabal DD, Song Z, Potian JG, Beuve A, McArdle JJ, Routh VH. Glucose, insulin, and leptin signaling pathways modulate nitric oxide synthesis in glucose-inhibited neurons in the ventromedial hypothalamus. Am J Physiol-Regul Integr Comp Physiol 2007; 292: R Hardie DG, Carling D. The AMP-activated protein kinase. Fuel gauge of the mammalian cell? Eur J Biochem 1997; 246: Turnley AM, Stapleton D, Mann RJ, Witters LA, Kemp BE, Bartlett PF. Cellular distribution and developmental expression of AMP-activated protein kinase isoforms in mouse central nervous system. J Neurochem 1999; 72: Mountjoy PD, Bailey SJ, Rutter GA. Inhibition by glucose or leptin of hypothalamic neurons expressing neuropeptide Y requires changes in AMP-activated protein kinase activity. Diabetologia 2007; 50: Fioramonti X, Contie S, Song Z, Routh VH, Lorsignol A, Penicaud L. Characterization of glucose-sensing neuron subpopulations in the arcuate nucleus. Diabetes 2007; 56: McCrimmon RJ, Fan X, Ding Y, Zhu W, Jacob RJ, Sherwin RS. Potential role for AMP-activated protein kinase in hypoglycemia sensing in the ventromedial hypothalamus. Diabetes 2004; 53: McCrimmon RJ, Fan X, Cheng H, McNay E, Chan O, Shaw M et al. Activation of AMP-activated protein kinase within the ventromedial hypothalamus amplifies counterregulatory hormone responses in rats with defective counterregulation. Diabetes 2006; 55: McCrimmon RJ, Shaw M, Fan X, Cheng H, Ding Y, Wang A et al. AMP-activated protein kinase (AMPK): a key mediator of hypoglycemia-sensing in the ventromedial hypothalamus (VMH). Diabetes 2006; 55 (Suppl. 1): A Claret M, Smith MA, Batterham RL, Selman C, Choudhury AI, Fryer LGD et al. AMPK is essential for energy homeostasis regulation and glucose sensing by POMC and AgRP neurons. J Clin Invest 2007; 117: Chan O, Zhu W, Ding Y, McCrimmon RJ, Sherwin RS. Blockade of GABA(A) receptors in the ventromedial hypothalamus further stimulates glucagon and sympathoadrenal but not the hypothalamopituitary-adrenal response to hypoglycemia. Diabetes 2006; 55: Mokan M, Mitrakou A, Veneman T, Ryan C, Korytkowski M, Cryer P et al. Hypoglycemia unawareness in IDDM. Diabetes Care 1994; 17: White NH, Skor DA, Cryer PE, Levandoski LA, Bier DM, Santiago JV. Identification of Type 1 diabetic patients at increased risk for hypoglycemia during intensive insulin therapy. New Engl J Med 1983; 308: Fukuda M, Tanaka A, Tahara Y, Ikegami H, Yamamoto Y, Kumahara Y et al. Correlation between minimal secretory capacity of pancreatic beta-cells and stability of diabetic control. Diabetes 1988; 37: Asplin CM, Paquette TL, Palmer JP. In vivo inhibition of glucagon secretion by paracrine beta cell activity in man. J Clin Invest 1981; 68: Maruyama H, Hisatomi A, Orci L, Grodsky GM, Unger RH. Insulin within islets is a physiologic glucagon release inhibitor. J Clin Invest 1984; 74: Samols E, Stagner JI, Ewart RB, Marks V. The order of islet microvascular cellular perfusion is B----A----D in the perfused rat pancreas. J Clin Invest 1988; 82: Zhou H, Tran PO, Yang S, Zhang T, LeRoy E, Oseid E et al. Regulation of alpha-cell function by the beta-cell during hypoglycemia in Wistar rats: the switch-off hypothesis. Diabetes 2004; 53: Zhou H, Zhang T, Oseid E, Harmon J, Tonooka N, Robertson RP. Reversal of defective glucagon responses to hypoglycemia in insulindependent autoimmune diabetic BB rats. Endocrinology 2007; 148: McCrimmon RJ, Evans ML, Jacob RJ, Fan X, Zhu Y, Shulman GI et al. AICAR and phlorizin reverse the hypoglycemia-specific defect in glucagon secretion in the diabetic BB rat. Am J Physiol-Endocrinol Metab 2002; 283: E Taborsky GJ Jr, Ahren B, Havel PJ. Autonomic mediation of glucagon secretion during hypoglycemia: implications for impaired alpha-cell responses in Type 1 diabetes. Diabetes 1998; 47: Group TDR. Epidemiology of severe hypoglycemia in the Diabetes Control and Complications Trial. American Journal of Medicine 1991; 90: Amiel SA, Sherwin RS, Simonson DC, Tamborlane WV. Effect of intensive insulin therapy on glycemic thresholds for counterregulatory hormone release. Diabetes 1988; 37: Davis MR, Shamoon H. Counterregulatory adaptation to recurrent hypoglycemia in normal humans. J Clin Endocrinol Metab 1991; 73: Davis SN, Galassetti P, Wasserman DH, Tate D. Effects of antecedent hypoglycemia on subsequent counterregulatory responses to exercise. Diabetes 2000; 49: Heller SR, Cryer PE. Reduced neuroendocrine and symptomatic responses to subsequent hypoglycemia after 1 episode of hypoglycemia in nondiabetic humans. Diabetes 1991; 40: Flanagan DE, Keshavarz T, Evans ML, Flanagan S, Fan X, Jacob RJ et al. Role of corticotrophin-releasing hormone in the impairment of counterregulatory responses to hypoglycemia. Diabetes 2003; 52: Powell AM, Sherwin RS, Shulman GI. Impaired hormonal responses to hypoglycemia in spontaneously diabetic and recurrently hypoglycemic rats. Reversibility and stimulus specificity of the deficits. J Clin Invest 1993; 92: Borg MA, Borg WP, Tamborlane WV, Brines ML, Shulman GI, Sherwin RS. Chronic hypoglycemia and diabetes impair counterregulation induced by localized 2-deoxy-glucose perfusion of the ventromedial hypothalamus in rats. Diabetes 1999; 48: Song Z, Routh VH. Recurrent hypoglycemia reduces the glucose sensitivity of glucose-inhibited neurons in the ventromedial hypothalamus nucleus. Am J Physiol-Regul Integr Comp Physiol 2006; 291: R Journal compilation 2008 Blackwell Publishing Ltd. Diabetic Medicine, 25,