Hindbrain catecholamine neurons control multiple glucoregulatory responses

Physiology & Behavior 89 (2006) 490 – 500

Hindbrain catecholamine neurons control multiple glucoregulatory responses Sue Ritter ⁎, Thu T. Dinh, Ai-Jun Li Programs in Neuroscience, Washington State University, Pullman, WA 99164-6520, USA Received 6 January 2006; received in revised form 5 May 2006; accepted 25 May 2006

Abstract Reduced brain glucose availability evokes an integrated constellation of responses that protect and restore the brain's glucose supply. These include increased food intake, adrenal medullary secretion, corticosterone secretion and suppression of estrous cycles. Our research has focused on mechanisms and neural circuitry underlying these systemic glucoregulatory responses. Using microinjection techniques, we found that localized glucoprivation of hindbrain but not hypothalamic sites, elicited key glucoregulatory responses, indicating that glucoreceptor cells controlling these responses are located in the hindbrain. Selective destruction of hindbrain catecholamine neurons using the retrogradely transported immunotoxin, anti-dopamine beta-hydroxylase conjugated to saporin (DSAP), revealed that spinally-projecting epinephrine (E) or norepinephrine (NE) neurons are required for the adrenal medullary response to glucoprivation, while E/NE neurons with hypothalamic projections are required for feeding, corticosterone and reproductive responses. We also found that E/NE neurons are required for both consummatory and appetitive phases of glucoprivic feeding, suggesting that multilevel collateral projections of these neurons coordinate various components of the behavioral response. Epinephrine or NE neurons co-expressing neuropeptide Y (NPY) may be the neuronal phenotype required for glucoprivic feeding: they increase NPY mRNA expression in response to glucoprivation and are nearly eliminated by DSAP injections that abolish glucoprivic feeding. In contrast, lesion of arcuate nucleus NPY neurons, using the toxin, NPY–saporin, does not impair glucoprivic feeding or hyperglycemic responses. Thus, hindbrain E/NE neurons orchestrate multiple concurrent glucoregulatory responses. Specific catecholamine phenotypes may mediate the individual components of the overall response. Glucoreceptive control of these neurons resides within the hindbrain. © 2006 Elsevier Inc. All rights reserved. Keywords: Glucoprivation; Norepinephrine neurons; Epinephrine neurons; Neuropeptide Y; Anti-dopamine beta-hydroxylase saporin; Hindbrain; NPY–saporin; Glucose regulation; Glucoreceptors

A continuous supply of glucose is essential for the function and survival of the brain. Increased food intake is one of a highly integrated constellation of responses evoked by reduced brain glucose availability. Other key responses to glucose deficit include adrenal medullary secretion, corticosterone secretion, glucagon secretion, and suppression of reproductive function. Together these glucoregulatory responses serve to restore, protect and maintain the availability of the brain's essential metabolic fuel. Our work has focused on the elucidation of the mechanisms and neural circuitry through which systemic glucoregulatory responses are elicited. Recently, our results have demonstrated that hindbrain catecholamine neurons are required components of

⁎ Corresponding author. Tel.: +1 509 335 8113; fax: +1 509 335 4650. E-mail address: [email protected] (S. Ritter). 0031-9384/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.physbeh.2006.05.036

this circuitry. The present review will focus on these results, with an emphasis on the glucoprivic feeding response. The stimulatory effect of norepinephrine (NE) on food intake has been known for nearly 30 years [1–3]. Early pharmacological experiments suggested that central NE or epinephrine (E) neurons are importantly involved in glucoprivic feeding [4,5]. Destruction of brain catecholamine systems, using intracerebroventricular 6hydroxydopamine (6HD), impairs glucoprivic feeding [6]. Unfortunately, 6HD, applied in this manner, damages both nigrostriatal and mesolimbic dopamine systems, and thereby produces generalized motor and motivational impairments [7]. Since these impairments may be sufficient to account for the reported deficits in glucoprivic feeding, 6HD lesions have not clarified the role of catecholamine neurons in this response. Biochemical studies showing increased NE turnover [8,9] or release [10] in response to systemic glucoprivation also implicate NE

S. Ritter et al. / Physiology & Behavior 89 (2006) 490–500

491

neurons, but do not provide insights into the nature of their role. Hindbrain mapping of glucoprivation-induced Fos-immunoreactivity has revealed that specific hindbrain E and NE cell groups are preferentially and selectively activated by glucoprivation [11]. Moreover, the fact that glucoprivically activated E/NE cell groups lie within the glucoreceptive zones identified in our cannulation mapping experiments [12], aroused further speculation that particular hindbrain E/NE neurons were important participants in glucoregulatory responses to glucoprivation. The recent availability of a highly selective retrogradely transported immunotoxin has made it possible to rigorously test the hypothesis that specific hindbrain E/NE neurons mediate responses to glucoprivation. 1. Anti-dopamine β-hydroxylase saporin (DSAP) selectively lesions brain NE and E neurons Anti-dopamine β-hydroxylase conjugated to saporin, which we refer to as DSAP, is an immunotoxin that can be used to selectively destroy NE and E neurons [13–16]. It consists of a monoclonal antibody against dopamine β-hydroxylase (DBH), conjugated to the ribosomal toxin, saporin (SAP). The DSAP molecule is selectively internalized by E and NE neurons due to their unique expression of the catecholamine biosynthetic enzyme, DBH, which is exposed to the extracellular space during synaptic activity. One of the interesting features of DSAP is that, unlike many other saporin conjugates, it is retrogradely transported [17,18]. From a technical standpoint, this feature of DSAP provides yet another avenue for selectivity, since we found that it can be injected into particular terminal areas to lesion only those neurons projecting to or through that site. The ability to be retrogradely transported is essential for DSAP toxicity when it is

Fig. 1. Diagram of rat hindbrain in sagittal plane showing approximate locations of noradrenergic (A) and adrenergic (C) cell groups. Neurons with projections to the spinal cord are present in A5, A6 and subcoeruleus, A7, C1 (rostral part), C2 and C3. Neurons with projections to the forebrain are present in A1, A2, A5, A6, C1 (middle and caudal parts), C2 and C3. Catecholamine neurons also innervate hindbrain sites.

Fig. 2. Coronal hindbrain sections showing catecholamine neurons of cell groups A1/C1 and C1 (middle portion) revealed by tyrosine hydroxylase immunoreactivity. These cell groups project to and heavily innervate the paraventricular nucleus of the hypothalamus (PVH), but do not project spinally. Sections shown in the top row are from the control rats; those in the middle row are from rats injected intraspinally with DSAP; sections in the bottom row are from rats injected with DSAP into the PVH. Neurons in these cell groups were not damaged by intraspinal DSAP, but were virtually eliminated by PVH injections [18].

injected into a catecholamine terminal area because the saporin must access the soma where ribosomes are concentrated in order to exert its toxic action [14]. We have exploited this property of DSAP in our study of the role of hindbrain catecholamine cell groups in glucoregulation. The organization of these cell groups, as shown by early immunohistochemical studies [19,20], is shown diagrammatically in Fig. 1. We nanoinjected DSAP into E/NE terminal areas in the paraventricular nucleus of the hypothalamus (PVH) [18,21,22], arcuate nucleus of the hypothalamus [23] and spinal cord [18]. We found DSAP toxicity to be highly reproducible and selective for the E/NE neurons known to innervate the injection site. As shown in Figs. 2 and 3, injection of DSAP into the PVH

Fig. 3. Coronal hypothalamic sections show NE and E terminals in the PVH revealed by DBH immunoreactivity in rats injected into the PVH with SAP control solution (left) or DSAP (right). DSAP, but not SAP, caused a profound reduction of DBH-immunoreactive terminals in the PVH [18]. 3V = third ventricle.

492

S. Ritter et al. / Physiology & Behavior 89 (2006) 490–500

induced by 2-deoxy-glucose (2DG), a glycolytic inhibitor [25] (Fig. 4), but did not impair the adrenal medullary hyperglycemic response to 2DG. The feeding deficit was highly selective, leaving deprivation-induced feeding and feeding in response to beta mercaptoacetate (MA)-induced lipoprivation intact. Injection into the spinal cord did not impair glucoprivic feeding, but severely reduced the hyperglycemic response (Fig. 5), and associated adrenal medullary secretion (Fig. 6) and Fos expression. In subsequent work [21], we found that PVH DSAP injection produced profound reduction of glucoprivation-induced corticosterone secretion, without impairing corticosterone secretion in response to forced swimming (Fig. 7) or the basal circadian rhythm of plasma corticosterone. We also found that DSAP impaired glucoprivation-induced suppression of estrus and lengthening of the estrous cycle, without altering basal estrous cycle length under basal conditions [22] (Fig. 8). 3. Hindbrain catecholamine neurons are necessary for the consummatory component of glucoprivic feeding

Fig. 4. Intake of pelleted rodent chow in response to metabolic challenges in rats previously injected with SAP or DSAP into the PVH. The top graph shows responses to subcutaneous injections of saline control (0.9% saline, 1 ml/kg), 2deoxy-D-glucose (2DG, 200 mg/kg), or a hypoglycemic dose of insulin (2 U/ kg). The bottom graph shows responses to intraperitoneal injections of saline or beta mercaptoacetate (MA, 68 mg/kg) or to overnight (18 h) food deprivation. DSAP selectively eliminated feeding responses to 2DG and hypoglycemia, without impairing responses to MA or deprivation [32].

destroys hindbrain NE and E neurons with projections to the PVH, as well as the terminals of those neurons in the PVH and medial hypothalamus. Results of our various experiments with PVH or arcuate nucleus microinjections of DSAP [18,21–23] show that DSAP did not lesion dopaminergic neurons or neurons containing corticotropin releasing hormone (CRH), oxytocin, vasopressin, neuropeptide Y (NPY) mRNA or Agouti Gene Related Protein (AGRP) mRNA in the vicinity of the injection site. When injected intraspinally into the vicinity of the intermediolateral cell column, DSAP did not destroy cholinergic neurons in the injection field. In addition, we found that DSAP-injected rats are healthy and clinically indistinguishable from normal controls or controls injected with unconjugated saporin (SAP), except that PVH injected DSAP is associated with a slowly-developing obesity. These findings have been summarized in a recent review [24]. 2. DSAP-induced selective destruction of hindbrain catecholamine neurons abolishes specific glucoregulatory responses In our initial studies [18], we found that PVH injections of DSAP abolished the feeding response to systemic glucoprivation

Appetitive behaviors and consummatory responses are distinct but closely integrated components of ingestion [26,27]. Appetitive responses include complex motivated behaviors involved in searching for and ingesting food. Consummatory responses include reflexes necessary for accepting, chewing and swallowing food once it is in the mouth. Results from experiments using decerebrate rats [28] have shown that the forebrain is not required for glucoprivation-induced enhancement of consummatory responses, leading to the view that these two phases of feeding are governed independently. Hindbrain catecholamine neurons provide dense innervation of forebrain areas implicated in appetitive behaviors, but also innervate hindbrain sites involved in visceral and oropharyngeal reflexes [29–31]. Thus, the distribution of NE and E terminals is compatible with a role for these neurons in coordinating both the appetitive and consummatory components of ingestion. We used DSAP to test the role of hindbrain E/NE neurons in consummatory responses to glucoprivation. It is possible, we thought, that DSAP lesions impair the feeding response by disconnecting hindbrain glucoreceptors from forebrain appetitive systems, while leaving the control of consummatory responses

Fig. 5. Blood glucose responses after subcutaneous injection at time 0 of saline (0.9%) or 2DG (200 mg/kg) in rats previously injected intraspinally with DSAP or control solution [18]. Intraspinal DSAP severely impaired the hyperglycemic response to systemic glucoprivation.

S. Ritter et al. / Physiology & Behavior 89 (2006) 490–500

493

Fig. 6. Effect of intraspinal SAP and DSAP on plasma catecholamine responses to insulin-induced hypoglycemia (top) and 5 min of forced swim [24]. Results confirm previous demonstrations that glucoprivation is a selective stimulus for adrenal medullary E secretion [94,95]. Swim stress, while not stimulating adrenal medullary secretion of E, did increase plasma NE. The latter indicates that spinal systems (possibly descending noncatecholaminergic neurons) controlling sympathetic postganglionic responses remained competent in the DSAP-lesioned rat to elicit sympathetic activation.

intact. We reasoned that the enhancement of consummatory responses by glucoprivation might still be demonstrable if these responses were catecholamine-independent or if they required catecholamine neurons not destroyed by the PVH DSAP injection. Rats nanoinjected with PVH DSAP or SAP were subsequently tested for consummatory responses to several challenges, using a dilute (40%) solution of lactose free milk delivered intraorally through chronic cheek fistulas as the test diet [32]. We found that DSAP-injected rats did not increase their consumption of milk delivered via the cheek fistula in response to either insulininduced hypoglycemia or systemic 2DG administration. However, the same rats did vigorously increase their consumption in response to either MA-induced blockade of fatty acid oxidation or to overnight food deprivation (Fig. 9). These results strongly suggest that both appetitive and consummatory responses to glucoprivation are controlled and coordinated by multilevel terminations of the same catecholamine neurons. They also leave open the possibility that the catecholamine neurons themselves are glucoreceptive.

responses. This is because the retrograde lesion destroys not only those projections to or through the injection site, but also collateral processes that emanate from the same affected cell bodies but innervate different sites.

4. Hindbrain catecholamine neurons are functionally specialized The fact that PVH DSAP injections impair or abolish glucoregulatory feeding, reproductive, adrenal medullary and corticosterone responses does not necessarily suggest that the same neurons mediate all of these responses, since functionally distinct neurons may have overlapping terminal beds and therefore be simultaneously lesioned by the DSAP injection. Moreover, the fact that PVH DSAP injections impair these responses does not necessarily indicate that catecholamine terminals in the PVH and surrounding diffusion zone are those most critical for the impaired

Fig. 7. Plasma corticosterone responses of PVH SAP or DSAP-injected rats during saline baseline conditions or insulin-induced hypoglycemia (top) and in response to forced swimming (bottom). The corticosterone response to hypoglycemia, but not forced swimming, was impaired by PVH DSAP injections [21].

494

S. Ritter et al. / Physiology & Behavior 89 (2006) 490–500

Another point worth considering is that DSAP lesions impair or abolish glucoregulatory feeding, reproductive, adrenal medullary and corticosterone responses without impairing the basal functioning of the same systems. This suggests the interesting possibility that hindbrain catecholamine involvement in these particular systems exclusively subserve the glucoprivic control. Even if this is the case, however, this does not suggest that glucoregulation is the only function of hindbrain catecholamine neurons. Catecholamine neurons are also known to be involved in responses to other physiological challenges, such as hypovolemia and hypotension [33–38] and in mediating some effects of cholecystokinin [39], but it seems likely that different catecholamine neurons are involved in these responses. This conclusion is strongly supported by the literature. For example, these and other studies show that catecholamine neurons important for sympathetic activation in response to hypotension are located in the retrofacial (rostral) portion of C1 and project spinally, while those that mediate the vasopressin response to hypovolemia are located in the A1 cell group and project to the PVH. We recently reported that hypoglycemia did not stimulate vasopressin secretion in intact rats, although corticosterone and food intake were stimulated by the same insulin dose, indicating that responses to hypovolemia and hypoglycemia are mediated by different catecholamine neurons under different sensory controls [40]. In the latter case, both functionally distinct groups project to the PVH. Finally, direct injection of DSAP into the NTS decreases the suppression of food intake in response to cholecystokinin [39]. Since in our study we found no deficit in the CCK response after PVH injection [18], the differing results are likely to be due to destruction of different populations of catecholamine neurons by the two injection strategies. Injection of DSAP into the PVH does not completely lesion the A2 cell group (since not all A2 neurons project to the PVH), but consistently reduces A2 cell numbers by approximately 50%. A local A2 injection of DSAP would be expected to lesion a different population of A2 neurons, one that includes those that do not project to the PVH and which may be functionally distinct. Thus, the differing findings with cholecystokinin may be a case in point, illustrating functional differences between individual catecholamine neurons. As the data continue to be generated, we expect that subgroups of hindbrain cate-

Fig. 9. Consumption of 40% milk solution (40% lactose free whole milk in tap water) delivered intraorally through a chronic cheek fistula to PVH SAP and DSAP-injected rats. Top: consumption after injection of saline control (0.9%), 2DG (200 mg/kg, sc), or insulin (2.0 U/kg, sc). Bottom: saline, mercaptoacetate (MA, 68 mg/kg, i.p.) or overnight (18 h) food deprivation. Consummatory responding to glucoprivation was selectively impaired by DSAP. Responses to deprivation and MA were not impaired [32].

cholamine neurons with distinct sensory inputs, neural projections and functions will be identified, even though these functioally distinct subpopulations may have overlapping terminal fields or co-mingled cell bodies. Teasing apart these functionally distinct catecholamine subpopulations is an important goal for future research. We, of course, are particularly interested in precise identification of those that are required for the various glucoregulatory responses. Fig. 8. Estrous cycles in PVH SAP and DSAP-injected rats under control conditions and in response to chronic glucoprivation (3 days of 2DG treatment). DSAP rats had estrous cycles that did not differ in length from SAP rats under control conditions. In response to chronic glucoprivation, estrous cycle length was significantly increased in SAP-treated, but not in DSAP-treated, rats [22].

5. Hindbrain NPY neurons participate in glucoregulatory responses One way to explore the division of labor between different populations of hindbrain catecholamine neurons is to isolate the

S. Ritter et al. / Physiology & Behavior 89 (2006) 490–500

functions of specific catecholamine subphenotypes, many of which have been identified already on the basis of co-expressed peptides and membrane receptors. We have made some progress in this regard by investigating the contribution to glucoprivic feeding of hindbrain catecholamine neurons that co-express NPY. Several lines of evidence suggest that NPY neurons are important for glucoprivic feeding. Injection of NPY antibodies into the PVH area reduces glucoprivic feeding [41] and knockout of the NPY gene reduces responsiveness to glucoprivation [42]. Although NPY neurons are found in various locations within the brain, our work indicates that hindbrain catecholamine/NPY co-expressing neurons are mediators of the glucoprivic feeding response. Glucoprivation increases both DBH [43] and NPY mRNA expression in these neurons [44] (Fig. 10). Due to their location, we can infer from our previous Fos mapping studies that glucoprivation produces a robust Fos response in this phenotypic

495

subgroup [11]. In addition, catecholamine/NPY neurons are virtually eliminated by PVH DSAP injections that abolish glucoprivic feeding [18,44]. Neuropeptide Y neurons are also found in the arcuate nucleus of the hypothalamus. These NPY neurons co-express AGRP. Expression of both NPY and AGRP mRNAs is increased by glucoprivation [45]. However, these increases in gene expression are abolished by injection of DSAP into the arcuate nucleus, which eliminates nearly all innervation to the arcuate nucleus by hindbrain catecholamine/NPY neurons [46]. This suggests that NPY/AGRP neurons do not respond directly to glucose deficit, but that their activation during glucoprivation is dependent on NPY/catecholamine projections from the hindbrain. If so, this raises the possibility that arcuate NPY/AGRP neurons are downstream of the hindbrain NPY/catecholamine neurons in a pathway involved in glucoprivic feeding.

Fig. 10. Hindbrain sections showing NPY mRNA hybridization signal in NE and E cell groups (top to bottom) A1, A1/C1, C1m (middle), C1r, C3. Rats were injected previously into the PVH with SAP or DSAP and treated 90 min prior to euthanasia with saline or 2DG (left to right). Cresyl violet counterstaining allowed visualization of cell bodies. Dark grains represent NPY mRNA hybridization signal. In SAP-treated controls, gene expression was increased in these cell groups, which provide the major NPY innervation of the hypothalamus from the hindbrain. DSAP lesions eliminated NPY mRNA-expressing neurons in these hindbrain areas. Bar = 5 um [44].

496

S. Ritter et al. / Physiology & Behavior 89 (2006) 490–500

Fig. 11. Drawing of coronal section of basomedial hypothalamus (−2.8 mm caudal to bregma) showing approximate area (oval) within which NPY, AGRP and CART mRNAs were quantified by densitometric analysis. Bilateral nanoinjections of NPY–SAP (48 ug in 100 nl) or blank-SAP (B-SAP) control were made at this level just dorsal to the arcuate nucleus. Quantification was done at 3 levels (−2.8± 0.5 mm from bregma). Bars show quantified hybridization signal at these levels for NPY, AGRP and CART from NPY–SAP rats, expressed as percent of B-SAP [47].

reduced in the NPY–SAP-injected rats. Dopamine β-hydroxylase and NPY terminals (presumably from the hindbrain) remained abundant in the NPY–SAP-injected area, despite the loss of hypothalamic NPY neuron cell bodies. Hypothalamic injection of NPY–SAP increased body weight, and eliminated weight loss and suppression of food intake induced by central leptin injection, as well as the stimulation of feeding by central ghrelin injection. Because both leptin and ghrelin are thought to produce their primary effects on feeding by actions on POMC/CART and NPY/ AGRP neurons within the arcuate nucleus [48–53], these deficits provide functional verification, consistent with the histochemical data, of the effectiveness of the NPY–SAP in destroying arcuate NPY/AGRP and POMC/CART neurons. Despite the effectiveness of the lesion, neither glucoprivic feeding nor the adrenal medullary hyperglycemic response was impaired in the same NPY–SAP-treated animals, (Fig. 13), indicating that hypothalamic NPY neurons are not required for these glucoregulatory responses. In short, DSAP injections that destroy hindbrain catecholamine/NPY neurons, but leave the arcuate NPY neurons intact, abolish glucoprivic feeding. Arcuate NPY–SAP injections that destroy arcuate NPY/AgRP neurons, but leave the hindbrain catecholamine/NPY neurons intact, do not impair glucoprivic feeding. Thus, arcuate NPY/AGRP neurons are not required for glucoprivic feeding. The functions associated with their activation during glucoprivation remain to be determined. 6. Glucoreceptor cells controlling systemic glucoregulatory responses are located in the hindbrain

We used another saporin conjugate, NPY–SAP, to test the importance of arcuate NPY/AGRP neurons in glucoprivic feeding [47]. This conjugate binds to NPY receptors [47], and presumably enters the cell by agonist-driven receptor internalization. Unlike DSAP, NPY–SAP did not destroy hindbrain catecholamine/NPY neurons or their hypothalamic terminals, but did destroy NPYreceptor expressing neurons in the arcuate nucleus, including NPY/AGRP and pro-opiomelanocortin (POMC)/cocaine and amphetamine related transcript (CART) neurons. There was a virtually complete loss of NPY, AGRP and CART mRNA expression in the basomedial hypothalamus, as shown in the quantified data presented in Fig. 11. In addition, NPY-Y1 receptor immunoreactivity in the arcuate nucleus (Fig. 12) was profoundly

Fig. 12. Coronal sections through the arcuate nucleus of rats injected into that nucleus with B-SAP control or NPY–SAP. Sections are stained to reveal NPY-Y1 receptor immunoreactivity. B-SAP-injected rats had significant levels of NPY-Y1 immunoreactivity, but this signal was eliminated by NPY–SAP, which binds to NPY receptors and lesion cells by NPY receptor-mediated internalization [47].

Receptor cells that detect glucose deficit and elicit systemic glucoregulatory responses are either the catecholamine neurons

Fig. 13. Food intake (top) and plasma glucose responses (bottom) to glucoprivation in rats injected into the arcuate nucleus with B-SAP control solution or NPY–SAP. Feeding responses were tested across a range of 2DG doses (0, 100, 200 and 400 mg/ kg). Plasma glucose was measured after a single 2DG dose (200 mg/kg). Feeding and glucose responses did not differ between B-SAP and NPY–SAP rats [47].

S. Ritter et al. / Physiology & Behavior 89 (2006) 490–500

themselves or are located in close proximity to them within the hindbrain. The most compelling evidence for this is the following: firstly, glucoprivic feeding and adrenal medullary responses are retained in decerebrate rats [28,54]; secondly, acute cerebral aqueduct occlusion blocks both feeding and adrenal medullary responses to lateral ventricular but not 4th ventricular injection of the glucoprivic agent, 5-thioglucose (5TG), indicating that 5TG must reach the hindbrain in order to elicit these responses [55]; and thirdly, cannula mapping studies have demonstrated hindbrain [12], but not hypothalamic or extrahypothalamic forebrain tissue sites [56,57], where localized glucoprivation elicits glucoregulatory responses. In addition to eliciting feeding and adrenal medullary responses to glucoprivation (Fig. 14) [12], nanoliter injections of 5TG into hindbrain sites, also elicits adrenal cortical and glucagon responses [58]. Work in other labs has shown that receptors controlling gonadotropin releasing hormone [59] and growth hormone responses [60] to glucoprivation are also located in the hindbrain. The lateral and third ventricle are effective injection sites for elicitation of glucoregulatory responses by 2DG or 5TG, but the aqueduct occlusion studies cited above, indicate that the effectiveness of such injections is dependent on the diffusion of the glucoprivic agent to the hindbrain. Tissue sites in the forebrain are generally ineffective, as noted above. The results of Borg et al. are exceptions to the generally negative results in the hypothalamus. These investigators have reported that microdialysis of 2DG into the ventromedial hypothalamus elicits glucoregulatory responses [61,62] and that the dialysis of glucose or lactate into this area blocks glucoregulatory responses to systemic hypoglycemia [63,64]. However, given the proximity of the ventromedial

497

hypothalamic microdialysis site to the third ventricle, the large surface area of the dialysis probe (1–1.5 mm) and the parameters of the infusion (100 mmol/l for 60 min), the localization of effects to the hypothalamus in these experiments remains questionable. Although the diffusion of dialysate to neighboring tissue sites was assessed, ventricular diffusion to hindbrain glucoreceptive sites was not convincingly ruled out as the mechanism of action. In short, the role of the hypothalamus in glucose homeostasis remains to be determined. However, the range of possible roles has been expanded by recent work suggesting that glucose sensing neurons in this area may respond to elevated glucose by reducing the rate of hepatic glucose production [65,66]. Nevertheless, this result must still be treated with caution because localization of effect to the ventral hypothalamus was assumed in this study, but not specifically assessed. In addition, a possible role for elevated glucose in suppressing hypothalamic feeding circuits and modulating metabolism, as proposed previously [67], should not be overlooked. The transduction mechanism utilized to detect glucose deficit by glucoreceptor cells that elicit systemic glucoregulatory responses is not known, since the glucoreceptor cells themselves have not been specifically identified. Two candidates for such a mechanism, however, are the ATP-sensitive K+ [K(ATP)] channel [68,69] and glucokinase, the pancreatic islet cell glucose sensing molecule, which is also found in the brain [70–72]. Electrophysiological studies have shown that some basomedial hypothalamic neurons that change their firing rate in response to changes in glucose concentration (i.e., “glucose sensing” neurons) express the K(ATP) channel and that this channel contributes to their responsiveness to glucose [73]. Support for the participation of glucokinase in glucoregulatory function has come in part from results using alloxan, which reduces glucokinase activity and impairs responses to glucose deficit [74–82]. However, localization, mechanism of action and toxicity studies required to interpret alloxan's effects in the brain have not been definitive to date. It is important to note that neither the K(ATP) channel nor glucokinase expression is uniquely expressed in areas presently associated with systemic glucoregulatory functions [70,73,83,84]. The K(ATP) channel in particular has been associated with neurons of diverse function [e.g.,83,84], such that the presence of the K(ATP) channel in a particular site does not necessarily mean that the site is involved in systemic glucoregulatory function. Moreover, glucose sensing molecules may be involved in controls important for food intake and metabolism that are not directly related to systemic responses to glucose deficit, as discussed elsewhere in this volume. 7. Role of glucoprivation in the patterning of “daily meals”

Fig. 14. Drawings of representative sections from caudal medulla (−13.80 mm caudal to bregma, top) and hypothalamus (−2.12 mm caudal to bregma, bottom) showing sites where nanoinjections of the glucoprivic agent, 5-thioglucose (5TG, 24 lg in 200 nl), stimulated feeding responses (stars), and ineffective injection sites (circles). Effective sites were clustered in and limited to the caudal medulla and appeared to be co-extensive with catecholamine cell groups. In nearly all cases, sites effective in stimulating feeding were also effective in eliciting hyperglycemic responses [12]. Drawings of brainstem are based on Paxinos [96].

There is no question that glucoprivation, when present, is a powerful stimulatory control of food intake. It can be activated despite large adipose stores. It can be activated even during periods of sustained and strongly suppressed food intake caused by chronic administration of exogenous leptin [85] and is not reduced under these conditions. Such demonstrations indicate the primacy of the brain's glucose requirement, as compared to body fat stores, as a control of appetite. Moreover, studies of

498

S. Ritter et al. / Physiology & Behavior 89 (2006) 490–500

hypoglycemia unawareness, a complication of insulin therapy in diabetic patients in which systemic glucoregulatory responses are not elicited by glucose deficit [86–88], provide evidence that disabling this control can result in cognitive impairment and is potentially lethal. But do the neural substrates that respond to glucoprivation participate in the patterning of daily meals? Certainly glucoprivation itself does not appear to be a required mechanism for meal initiation, since during ad libitum feeding with unrestricted food availability meals are usually initiated under euglycemic conditions. Nevertheless, the total supply of glycogen stored in the body is very limited (only about 800 kcal in humans) [89], suggesting that under more natural conditions in which animals expend considerable energy foraging for food, wider variations in glucose availability may occur. Under these conditions glucose decrements may drive feeding behaviors to a greater degree than they appear to in typical laboratory environments. It is also important to consider that glucoregulatory responses prevent, as well as reverse, glucoprivic conditions. To be effective in this regard, they are elicited hierarchically beginning prior to a crisis, while glucose is still within the normal range. Evidence from human studies using a stepped hypoglycemic clamp technique [90,91] indicates that cessation of insulin secretion is the first line of defense against glucoprivation, occurring at blood glucose concentrations of about 4.6 mM (very close to the normal 5 mM glucose concentration). The suppression of insulin secretion preserves blood glucose without mobilizing glycogen. Glucagon and epinephrine secretion are more costly responses, elevating glucose by mobilizing and diminishing glycogen stores. Glucagon and adrenal medullary secretion are triggered at glycemic levels of about 3.8 mM, the approximate threshold for reduction of brain glucose uptake [92]. Glucocorticoid secretion increases at about 3.2 mM, still somewhat above the level at which autonomic signs and cognitive dysfunction occur (about 3.0 and 2.7 mM, respectively) [91]. Unfortunately, we do not have definitive data to suggest where elicitation of feeding fits into this hierarchy of responses. However, because food acquisition potentially requires extended energy expenditure, one line of logic would predict that the feeding response is triggered earlier and at a higher glucose level than adrenal medullary and glucagon secretion. This would enable the animals to forage for food prior to depletion of energy reserves. Alternatively, feeding and endocrine response thresholds might be the same, with the endocrine responses supporting the physical activity required for foraging. The possibility that feeding and endocrine controls of glycemia are elicited at different glucose thresholds raises a tantalizing issue. To date, most studies of glucoprivic feeding have been limited to situations in which glucoprivation is induced acutely by direct blockade of intracellular glucose metabolism or by acute hypoglycemia, hence the name “glucoprivic control”. Induction of acute glucoprivation, which provides a clear metabolic stimulus for feeding, has been a very useful and highly productive approach for the identification of the underlying neural substrate. However, the name “glucoprivic control” may have limited our concept of how the control might operate, confining its neural substrate to the service of this one particular stimulus. Acute ongoing glucoprivation may not be the only paradigm for

triggering the underlying mechanism for stimulation of feeding. The phenomenon of “delayed glucoprivic feeding” [93], in which the feeding response is temporally uncoupled from the glycemic evidence of ongoing glucoprivation, should encourage us to consider other ways in which glycemic history may contribute to food intake. Moreover, the same neural circuitry that responds to glucoprivation may respond to or be modulated by nonglucoprivic signals that alter its response properties. Hopefully, as we home in on the specific neural circuits and “glucoreceptor” cells controlling the glucoprivic feeding response, it will become possible to examine their participation in feeding, endocrine and metabolic controls under a variety of physiological circumstances and using an array of new techniques. 8. Summary Although catecholamine neuron participation in the control of food intake and autonomic and neuroendocrine responses has long been suspected, results of DSAP lesions have made it clear that hindbrain catecholamine neurons play specific and essential roles in eliciting a variety of responses to glucose deficit. The influence of these neurons extends across all levels of the neuroaxis to orchestrate responses of multiple complex systems. The degree to which individual hindbrain catecholamine neurons are anatomically devoted to specific glucoregulatory and nonglucoregulatory functions is an important area for further investigation. The isolation of glucoreceptor cells that control these catecholamine neurons and the identification of the specific sensory transduction mechanisms that equip receptor cells to detect glucose deficit will greatly advance this field. Progress in these areas will facilitate the prevention and treatment of hypoglycemia associated autonomic failure, a potentially fatal side effect of insulin therapy in diabetics in which the glucoregulatory response system is disabled, and will enhance our understanding of the contribution of the glucoprivic control to overall control of food intake and energy homeostasis. Acknowledgements This work was supported by PHS grants R01 DK 40498 and NS 4552004 and by the Juvenile Diabetes Research Foundation International and the American Diabetes Association. References [1] Ritter RC, Epstein AN. Control of meal size by central noradrenergic action. Proc Natl Acad Sci U S A 1975;72(9):3740–3. [2] Ritter S, Wise D, Stein L. Neurochemical regulation of feeding in the rat: facilitation by alpha-noradrenergic, but not dopaminergic, receptor stimulants. J Comp Physiol Psychol 1975;88(2):778–84. [3] Leibowitz SF. Central adrenergic receptors and the regulation of hunger and thirst. Res Publ Assoc Res Nerv Ment Dis 1972;50:327–58. [4] Booth DA. Modulation of the feeding response to peripheral insulin, 2deoxyglucose or 3-O-methyl glucose injection. Physiol Behav 1972;8 (6):1069–76. [5] Muller EE, Cocchi D, Mantegazza P. Brain adrenergic system in the feeding response induced by 2-deoxy-D-glucose. Am J Physiol 1972;223: 945–50.

S. Ritter et al. / Physiology & Behavior 89 (2006) 490–500 [6] Stricker EM, Friedman MI, Zigmond MJ. Glucoregulatory feeding by rats after intraventricular 6-hydroxydopamine or lateral hypothalamic lesions. Science 1975;189(4206):895–7. [7] Marshall JF, Richardson JS. Nigrostriatal bundle damage and the lateral hypothalamic syndrome. J Comp Physiol Psychol 1974;87:808–30. [8] Bellin SI, Ritter S. Insulin-induced elevation of hypothalamic norepinephrine turnover persists after glucorestoration unless feeding occurs. Brain Res 1981;217(2):327–37. [9] Rowland NE. Effects of glucose and fat antimetabolites on norepinephrine turnover in rat hypothalamus and brainstem. Brain Res 1992;595(2):291–4. [10] Beverly JL, De Vries MG, Bouman SD, Arseneau LM. Noradrenergic and GABAergic systems in the medial hypothalamus are activated during hypoglycemia. Am J Physiol Regul Integr Comp Physiol 2001;280(2): R563–9. [11] Ritter S, Llewellyn-Smith I, Dinh TT. Subgroups of hindbrain catecholamine neurons are selectively activated by 2-deoxy-D-glucose induced metabolic challenge. Brain Res 1998;805(1–2):41–54. [12] Ritter S, Dinh TT, Zhang Y. Localization of hindbrain glucoreceptive sites controlling food intake and blood glucose. Brain Res 2000;856(1–2):37–47. [13] Wrenn CC, Picklo MJ, Lappi DA, Robertson D, Wiley RG. Central noradrenergic lesioning using anti-DBH-saporin: anatomical findings. Brain Res 1996;740(1–2):175–84. [14] Wiley RG, Kline IR. Neuronal lesioning with axonally transported toxins. J Neurosci Methods 2000;103:73–82. [15] Picklo MJ, Wiley RG, Lappi DA, Robertson D. Noradrenergic lesioning with an anti-dopamine beta-hydroxylase immunotoxin. Brain Res 1994;666 (2):195–200. [16] Picklo MJ, Wiley RG, Lonce S, Lappi DA, Robertson D. Anti-dopamine betahydroxylase immunotoxin-induced sympathectomy in adult rats. J Pharmacol Exp Ther 1995;275(2):1003–10. [17] Blessing WW, Lappi DA, Wiley RG. Destruction of locus coeruleus neuronal perikarya after injection of anti-dopamine-B-hydroxylase immunotoxin into the olfactory bulb of the rat. Neurosci Lett 1998;243(1–3):85–8. [18] Ritter S, Bugarith K, Dinh TT. Immunotoxic destruction of distinct catecholamine subgroups produces selective impairment of glucoregulatory responses and neuronal activation. J Comp Neurol 2001;432(2):197–216. [19] Dahlstrom A, Fuxe K. Localization of monoamines in the lower brain stem. Experientia 1964;20(7):398–9. [20] Hillarp NA, Fuxe K, Dahlstrom A. Demonstration and mapping of central neurons containing dopamine, noradrenaline, and 5-hydroxytryptamine and their reactions to psychopharmaca. Pharmacol Rev 1966;18(1):727–41. [21] Ritter S, Watts AG, Dinh TT, Sanchez-Watts G, Pedrow C. Immunotoxin lesion of hypothalamically projecting norepinephrine and epinephrine neurons differentially affects circadian and stressor-stimulated corticosterone secretion. Endocrinology 2003;144(4):1357–67. [22] I'Anson H, Sundling LA, Roland SM, Ritter S. Immunotoxic destruction of distinct catecholaminergic neuron populations disrupts the reproductive response to glucoprivation in female rats. Endocrinology 2003;144: 4325–31. [23] Fraley GS, Ritter S. Immunolesion of norepinephrine and epinephrine afferents to medial hypothalamus alters basal and 2DG-induced NPY and AGRP mRNA expression in the arcuate nucleus. Endocrinology 2003;144 (1):75–83. [24] Ritter S, Dinh TT, Bugarith K, Salter DM. Chemical dissection of brain glucoregulatory circuitry. In: Wiley RG, Lappi DA, editors. Molecular neurosurgery with targeted toxins. Totowa, N.J.: Humana Press, Inc.; 2005. p. 181–218. [25] Brown J. Effects of 2-deoxy-D-glucose on carbohydrate metabolism: review of the literature and studies in the rat. Metabolism 1962;11:1098–112. [26] Grill HJ. Production and regulation of ingestive consummatory behavior in the chronic decerebrate rat. Brain Res Bull 1980;5:79–87. [27] Grill HJ, Kaplan JM. The neuroanatomical axis for control of energy balance. Front Neuroendocrinol 2002;23(1):2–40. [28] Flynn FW, Grill HJ. Insulin elicits ingestion in decerebrate rats. Science 1983;221(4606):188–90. [29] Cunningham Jr ET, Sawchenko PE. Dorsal medullary pathways subserving oromotor reflexes in the rat: implications for the central neural control of swallowing. J Comp Neurol 2000;417(4):448–66.

499

[30] Levitt P, Moore RY. Origin and organization of brainstem catecholamine innervation in the rat. J Comp Neurol 1979;186(4):505–28. [31] Mtui EP, Anwar M, Reis DJ, Ruggiero DA. Medullary visceral reflex circuits: local afferents to nucleus tractus solitarii synthesize catecholamines and project to thoracic spinal cord. J Comp Neurol 1995;351(1):5–26. [32] Hudson B, Ritter S. Hindbrain catecholamine neurons mediate consummatory responses to glucoprivation. Physiol Behav 2004;82(2–3):241–50. [33] Guyenet PG, Schreihofer AM, Stornetta RL. Regulation of sympathetic tone and arterial pressure by the rostral ventrolateral medulla after depletion of C1 cells in rats. Ann N Y Acad Sci 2001;940:259–69. [34] Guyenet PG, Stornetta RL. Inhibition of sympathetic preganglionic discharges by epinephrine and alpha-methylepinephrine. Brain Res 1982;235(2):271–83. [35] Madden CJ, Ito S, Rinaman L, Wiley RG, Sved AF. Lesions of the C1 catecholaminergic neurons of the ventrolateral medulla in rats using antiDbetaH-saporin. Am J Physiol 1999;277(4 Pt 2):R1063–75. [36] Day TA, Ferguson AV, Renaud LP. Noradrenergic afferents facilitate the activity of tuberoinfundibular neurons of the hypothalamic paraventricular nucleus. Neuroendocrinology 1985;41(1):17–22. [37] Sladek CD, Kapoor JR. Neurotransmitter/neuropeptide interactions in the regulation of neurohypophyseal hormone release. Exp Neurol 2001;171 (2):200–9. [38] Blessing WW. The lower brainstem and bodily homeostasis. New York: Oxford University Press; 1997. [39] Rinaman L. Hindbrain noradrenergic lesions attenuate anorexia and alter central cFos expression in rats after gastric viscerosensory stimulation. J Neurosci 2003;23(31):10084–92. [40] Dinh TT, Flynn FW, Ritter S. Hypovolemia and hypoglycemia activate different hindbrain catecholamine neurons with projections to the hypothalamus. Am J Physiol in press. [41] He B, White BD, Edwards GL, Martin RJ. Neuropeptide Yantibody attenuates 2-deoxy-D-glucose induced feeding in rats. Brain Res 1998;781(1–2):348–50. [42] Sindelar DK, Ste Marie L, Miura GI, Palmiter RD, McMinn JE, Morton GJ, et al. Neuropeptide Y is required for hyperphagic feeding in response to neuroglucopenia. Endocrinology 2004;145(7):3363–8. [43] Li A-J, Wang Q, Ritter S. Differential responsiveness of dopamine-betahydroxylase gene expression to glucoprivation in different catecholamine cell groups. Endocrinology In press. [44] Li AJ, Ritter S. Glucoprivation increases expression of neuropeptide Y mRNA in hindbrain neurons that innervate the hypothalamus. Eur J Neurosci 2004;19(8):2147–54. [45] Sergeyev V, Broberger C, Gorbatyuk O, Hokfelt T. Effect of 2-mercaptoacetate and 2-deoxy-D-glucose administration on the expression of NPY, AGRP, POMC, MCH and hypocretin/orexin in the rat hypothalamus. NeuroReport 2000;11(1):117–21. [46] Fraley GS, Ritter S. Immunolesion of norepinephrine and epinephrine afferents to medial hypothalamus alters basal and 2-deoxy-D-glucose-induced neuropeptide Y and agouti gene-related protein messenger ribonucleic acid expression inthe arcuate nucleus. Endocrinology 2003;144(1):75–83. [47] Bugarith K, Dinh TT, Li AJ, Speth RC, Ritter S. Basomedial hypothalamic injections of neuropeptide y conjugated to saporin selectively disrupt hypothalamic controls of food intake. Endocrinology 2005;146 (3):1179–91. [48] Cowley MA, Smart JL, Rubinstein M, Cerdan MG, Diano S, Horvath TL, et al. Leptin activates anorexigenic POMC neurons through a neural network in the arcuate nucleus. Nature 2001;411(6836):480–4. [49] Cone RD, Cowley MA, Butler AA, Fan W, Marks DL, Low MJ. The arcuate nucleus as a conduit for diverse signals relevant to energy homeostasis. Int J Obes Relat Metab Disord 2001;25(Suppl 5):S63–7. [50] Elias CF, Saper CB, Maratos-Flier E, Tritos NA, Lee C, Kelly J, et al. Chemically defined projections linking the mediobasal hypothalamus and the lateral hypothalamic area. J Comp Neurol 1998;402(4):442–59. [51] Elias CF, Lee C, Kelly J, Aschkenasi C, Ahima RS, Couceyro PR, et al. Leptin activates hypothalamic CART neurons projecting to the spinal cord. Neuron 1998;21(6):1375–85. [52] Elias CF, Aschkenasi C, Lee C, Kelly J, Ahima RS, Bjorbaek C, et al. Leptin differentially regulates NPY and POMC neurons projecting to the lateral hypothalamic area. Neuron 1999;23(4):775–86.

500

S. Ritter et al. / Physiology & Behavior 89 (2006) 490–500

[53] Elmquist JK, Ahima RS, Elias CF, Flier JS, Saper CB. Leptin activates distinct projections from the dorsomedial and ventromedial hypothalamic nuclei. Proc Natl Acad Sci U S A 1998;95(2):741–6. [54] DiRocco RJ, Grill HJ. The forebrain is not essential for sympathoadrenal hyperglycemic response to glucoprivation. Science 1979;204(4397): 1112–4. [55] Ritter RC, Slusser PG, Stone S. Glucoreceptors controlling feeding and blood glucose: location in the hindbrain. Science 1981;213(4506):451–2. [56] Miselis RR, Epstein AN. Feeding induced by intracerebroventricular 2deoxy-D-glucose in the rat. Am J Physiol 1975;229(5):1438–47. [57] Berthoud HR, Mogenson GJ. Ingestive behavior after intracerebral and intracerebroventricular infusions of glucose and 2-deoxy-D-glucose. Am J Physiol 1977;233(3):R127–33. [58] Andrew S, Pedrow C, Ritter S. Localized glucoprivation of hindbrain but not hypothalamic sites stimulates glucagon and corticosterone secretion. Abstr Soc Neurosci 2003 [Program number 237.8]. [59] Murahashi K, Bucholtz DC, Nagatani S, Tsukahara S, Tsukamura H, Foster DL, et al. Suppression of luteinizing hormone pulses by restriction of glucose availability is mediated by sensors in the brain stem. Endocrinology 1996;137(4):1171–6. [60] Penicaud L, Pajot MT, Thompson DA. Evidence that receptors controlling growth hormone and hyperglycemic responses to glucoprivation are located in the hindbrain. Endocr Res 1990;16(4):461–75. [61] Borg WP, Sherwin RS, During MJ, Borg MA, Shulman GI. Local ventromedial hypothalamus glucopenia triggers counterregulatory hormone release. Diabetes 1995;44(2):180–4. [62] 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(3):584–7. [63] 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 (2):361–5. [64] Borg MA, Tamborlane WV, Shulman GI, Sherwin RS. Local lactate perfusion of the ventromedial hypothalamus suppresses hypoglycemic counterregulation. Diabetes 2003;52(3):663–6. [65] Lam TK, Gutierrez-Juarez R, Pocai A, Rossetti L. Regulation of blood glucose by hypothalamic pyruvate metabolism. Science 2005;309(5736):943–7. [66] Pocai A, Lam TK, Gutierrez-Juarez R, Obici S, Schwartz GJ, Bryan J, et al. Hypothalamic K(ATP) channels control hepatic glucose production. Nature 2005;434(7036):1026–31. [67] Panksepp J, Rossi III J. D-glucose infusions into the basal ventromedial hypothalamus and feeding. Behav Brain Res 1981;3(3):381–92. [68] Routh VH. Glucose-sensing neurons: are they physiologically relevant? Physiol Behav 2002;76(3):403–13. [69] Levin BE, Dunn-Meynell AA, Routh VH. Brain glucosensing and the K (ATP) channel. Nat Neurosci 2001;4(5):459–60. [70] Li B, Xi X, Roane DS, Ryan DH, Martin RJ. Distribution of glucokinase, glucose transporter GLUT2, sulfonylurea receptor-1, glucagon-like peptide-1 receptor and neuropeptide Y messenger RNAs in rat brain by quantitative real time RT-PCR. Brain Res Mol Brain Res 2003;113(1–2):139–42. [71] Maekawa F, Toyoda Y, Torii N, Miwa I, Thompson RC, Foster DL, et al. Localization of glucokinase-like immunoreactivity in the rat lower brain stem: for possible location of brain glucose-sensing mechanisms. Endocrinology 2000;141(1):375–84. [72] Dunn-Meynell AA, Routh VH, Kang L, Gaspers L, Levin BE. Glucokinase is the likely mediator of glucosensing in both glucose-excited and glucoseinhibited central neurons. Diabetes 2002;51(7):2056–65. [73] 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(1–2):41–54.

[74] Woods SC, McKay LD. Intraventricular alloxan eliminates feeding elicited by 2-deoxyglucose. Science 1978;202(4373):1209–11. [75] Sanders NM, Dunn-Meynell AA, Levin BE. Third ventricular alloxan reversibly impairs glucose counterregulatory responses. Diabetes 2004;53 (5):1230–6. [76] Ritter S, Strang M. Fourth ventricular alloxan injection causes feeding but not hyperglycemia in rats. Brain Res 1982;249(1):198–201. [77] Murnane JM, Ritter S. Intraventricular alloxan impairs feeding to both central and systemic glucoprivation. Physiol Behav 1985;34(4):609–13. [78] Murnane JM, Ritter S. Alloxan-induced glucoprivic feeding deficits are blocked by D-glucose and amygdalin. Pharmacol Biochem Behav 1985;22 (3):407–13. [79] Ritter S, Murnane JM, Ladenheim EE. Glucoprivic feeding is impaired by lateral or fourth ventricular alloxan injection. Am J Physiol 1982;243(3): R312–7. [80] Kinoshita M, I'Anson H, Tsukamura H, Maeda K. Fourth ventricular alloxan injection suppresses pulsatile luteinizing hormone release in female rats. J Reprod Dev 2004;50(3):279–85. [81] Kang L, Routh VH, Kuzhikandathil EV, Gaspers LD, Levin BE. Physiological and molecular characteristics of rat hypothalamic ventromedial nucleus glucosensing neurons. Diabetes 2004;53(3):549–59. [82] Moriyama R, Tsukamura H, Kinoshita M, Okazaki H, Kato Y, Maeda K. In vitro increase in intracellular calcium concentrations induced by low or high extracellular glucose levels in ependymocytes and serotonergic neurons of the rat lower brainstem. Endocrinology 2004;145(5):2507–15. [83] Karschin C, Ecke C, Ashcroft FM, Karschin A. Overlapping distribution of K(ATP) channel-forming Kir6.2 subunit and the sulfonylurea receptor SUR1 in rodent brain. FEBS Lett 1997;401(1):59–64. [84] Mourre C, Widmann C, Lazdunski M. Sulfonylurea binding sites associated with ATP-regulated K+ channels in the central nervous system: autoradiographic analysis of their distribution and ontogenesis, and of their localization in mutant mice cerebellum. Brain Res 1990;519(1–2):29–43. [85] Wiater MF, Ritter S. Leptin does not attenuate the hyperphagia incuded by 2-deoxy-D-glucose (2DG). Ann N Y Acad Sci 1999;892:334. [86] Cryer PE. Hypoglycemia-associated autonomic failure in insulin-dependent diabetes mellitus. Adv Pharmacol 1998;42:620–2. [87] Sanders NM, Ritter S. Repeated 2-deoxy-D-glucose-induced glucoprivation attenuates Fos expression and glucoregulatory responses during subsequent glucoprivation. Diabetes 2000;49(11):1865–74. [88] Sanders NM, Ritter S. Acute 2DG-induced glucoprivation or dexamethasone abolishes 2DG-induced glucoregulatory responses to subsequent glucoprivation. Diabetes 2001;50(12):2831–6. [89] Bray GA. Appetite control in adults. In: Fernstrom JD, Miller GD, editors. Appetite and body weight regulation: sugar, fat, and macronutrient substitutes. Boca Raton: CRC Press; 1994. p. 17–31. [90] Cryer PE. Glucose counterregulation: prevention and correction of hypoglycemia in humans. Am J Physiol 1993;264(2(Pt 1)):P E149–55. [91] Schwartz NS, Clutter WE, Shah SD, Cryer PE. Glycemic thresholds for activation of glucose counterregulatory systems are higher than the threshold for symptoms. J Clin Invest 1987;79(3):777–81. [92] Siesjo BK. Utilization of substrates by brain tissues. Brain energy metabolism. New York: Wiley; 1978. p. 101–30. [93] Ritter RC, Roelke M, Neville M. Glucoprivic feeding behavior in absence of other signs of glucoprivation. Am J Physiol 1978;234:E617–21. [94] Vollmer RR, Balcita JJ, Sved AF, Edwards DJ. Adrenal epinephrine and norepinephrine release to hypoglycemia measured by microdialysis in conscious rats. Am J Physiol 1997;273(5 Pt 2):R1758–63. [95] Morrison SF, Cao WH. Different adrenal sympathetic preganglionic neurons regulate epinephrine and norepinephrine secretion. Am J Physiol Regul Integr Comp Physiol 2000;279(5):R1763–75. [96] Paxinos G, Watson C. The rat brain in stereotaxic coordinates. Third ed. San Diego: Academic Press; 1997.