INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

NEUROCHEMICAL EVENTS in the hypothalamus coordinate the activities of behavioral, endocrine, and autonomic systems to regulate basal metabolism. In the ventral hypothalamus, firing rates of specific neurons are influenced by glucose availability (22) and affect the activity of the sympathetic nervous system (26). The neural networks involved in integrating and maintaining nutrient homeostasis are unknown. In response to a glucoprivic challenge, induced by 2-deoxy-D-glucose (2-DG), concentrations of the neurotransmitter GABA increased in the ventromedial nucleus of the hypothalamus (VMN) and decreased in the lateral hypothalamic area (LHA) (2). In the same brain region, norepinephrine (NE) turnover was reported to be higher during glucoprivation (30, 31).

The activity of noradrenergic systems in the hypothalamus is likely to be involved in compensatory responses to glucoprivation. Circulating glucose and glucose-mobilizing hormones are increased in response to microinjection of NE into the medial hypothalamus (4, 32). Injection of NE into the VMN also stimulated food intake, as did the GABA_A receptor agonist muscimol (7, 12, 14). The increase in food intake by NE administration may be through GABA. Coadministering phentolamine, an -adrenergic antagonist, with muscimol did not block the feeding response to muscimol; however, the increased food intake after NE injection was blocked by coinjecting bicuculline, a GABA_A receptor antagonist (7). Thus part of the compensatory responses to a glucoprivic episode may involve NE mediation of hypothalamic GABA release in the VMN. In the hypothalamus, NE facilitated GABA release in a dose-dependent manner, an effect mimicked by clonidine, an ₂-adrenoceptor agonist, and blocked by yohimbine, an ₂-adrenoreceptor antagonist (16, 25).

The first objective of the present study was to characterize any change in noradrenergic activity in discrete hypothalamic areas during an acute period of glucoprivation. Reports of increased NE turnover in the hypothalamus during glucoprivation (30, 31) lacked temporal resolution, and the timing of increased NE release using push-pull (17, 18) or microdialysis (28) are in disagreement as to the timing of the increase. Microdialysis probes were used to monitor extracellular NE in the VMN, LHA, and paraventricular nucleus (PVN) of rats during a 2-DG-induced glucoprivic challenge. Because GABA and NE increased in the VMN in a similar pattern after 2-DG, the second objective of the present study was to evaluate the interaction between noradrenergic and GABAergic systems in the hypothalamus and the relationship between these neurotransmitters during an acute glucoprivic episode induced by 2-DG. Microdialysis probes placed into the VMN or LHA were used to monitor NE or GABA synaptic overflow in response to pharmacological manipulations. The results of the present study confirm the interaction of NE and GABA in the VMN during a glucoprivic episode. The increases in extracellular GABA concentration after 2-DG were mediated by an increase in NE. Furthermore, it appears that a different adrenoceptor subtype influences each peak in GABA activity.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

This study was approved by the Laboratory Animal Care Advisory Committee of the University of Illinois. Male Sprague-Dawley rats, ~250 g, were housed singly in Plexiglas cages (30 × 30 × 38 cm) in a light (12:12-h light-dark cycle; lights on at 0700)- and temperature (26 ± 2°C)-controlled room. Fresh water and rodent diet (Harlan Teklabs, Madison, WI) were available at all times, except during sample collection periods.

After a 1-wk acclimation period, rats were anesthetized with a mixture of ketamine HCl, xylazine HCl, and acepromazine (30:6:1 mg/kg im). The level of anesthesia was monitored and maintained at an appropriate level throughout the surgical procedure. The top of the head and the neck of each rat were shaved, and the skin was washed with povidone-iodine 10% (Betadine). A jugular vein catheter was aseptically placed through a vertical incision in the neck. The right external jugular vein was isolated, and a 4-cm segment of Silastic tubing (0.025 in. ID × 0.037 in. OD) was inserted toward the heart. The catheter was secured with 5-0 suture, tunneled under the skin, and exteriorized through an incision on top of the head before the skin in the neck was closed with wound staples. A piece of 21-gauge stainless steel tubing was inserted onto the end of the catheter, and the catheter was filled with a 40% polyvinylpyrrolidone solution containing 500 U heparin/ml and capped with a sealed piece of Tygon tubing to maintain patency. The rat was placed into a stereotaxic instrument (ASI Instruments, Warren, MI), and a guide cannula (23 gauge) was positioned 1 mm dorsal to the sample site using the stereotaxic atlas of Paxinos and Watson (23). Coordinates for sample sites were VMN [anteroposterior (AP) = 6.5, lateral (L) = 0.8, dorsal (D) = 9.6 below dura], LHA (AP = 6.5, L = 1.8, D = 8.8 below dura), or PVN (AP = 7.2, L = 0.6, D = 8.2 below dura). The guide cannula and end of venous catheters were fixed in position with dental acrylic cement and anchored to the skull with four stainless steel screws (Small Parts, Miami Beach, FL). After surgery, rats were monitored until they had completely recovered from the anesthetic. Postsurgical analgesia was provided by butorphenol (0.5 mg/kg sc; Torbuterol).

At the end of each experiment, rats were anesthetized and their hearts were exposed. The right atrium was punctured, and ~60 ml of chilled saline followed by ~60 ml of 10% Formalin solution were perfused through the brain via the left ventricle. The Formalin-fixed brain was removed from the skull, and intrahypothalamic cannula position was verified histologically.

Sample collection. Rats were allowed 7-10 days of recovery after surgery, and only animals with body weights greater than on the day of surgery were used. Blood and dialysate samples were collected via a weighted counterbalance lever with liquid swivel (Instech, Plymouth Meeting, PA) from unrestrained animals in their home cages during the mid-light phase. On the day of each experiment, food was removed from the rat's cage, and the in vitro recovery of concentric microdialysis probes [0.2 mm diameter × 1.0 mm cuprophan (Akzo Nobel Fraser, Wuppertal, Germany) membrane] was determined before probes were placed into the VMN, LHA, or PVN. Probe efficiencies were 3.5-6.6%. Probes were in place at least 3 h before samples were collected to allow for stable baseline neurotransmitter concentrations (2). Probes were continuously perfused with Krebs-Ringer buffer (KRB) (in mM: 147 NaCl, 4 KCl, 3.4 CaCl₂; pH 6.4) at 1.0 µl/min via a microinfusion pump (Bioanalytical Systems, West Lafayette, IN). Food was removed from cages when probes were placed into the rat brains and returned after the last sample was collected.

Sample analysis. Catecholamines were analyzed on a Dynamax SD-200 system (Varian Instruments, Woburn, MA) by reverse-phase HPLC and electrochemical detection. Samples (5-10 µl) were injected onto a 150 × 2 mm C₁₈ (3 µm) Hypersil column (Keystone Scientific, Bellfonte, PA) fitted with a 2-mm C₁₈ (3 µm) Hypersil javalin guard column (Keystone Scientific, Bellfonte, PA). Mobile phase (pH 3.0) was 75 mM NaH₂PO₄, 1.7 mM L-octanesulfonic acid, 25 µM Na₂EDTA, 10% (vol/vol) acetonitrile, and 0.1% (vol/vol) tetrahydrofuran. A DECADE electrochemical detector fitted with a VT-03 glassy carbon electrode (Antec Leyden, Leiden, The Netherlands) set at +0.75 V was used with Dynamax MacIntegrator II and "C" module programs (Rainin Instruments, Woburn, MA) for peak integration and quantification. With the use of this method, sensitivity limits were 200 fg.

Experiment 1: NE in VMH, LHA, and PVN after 2-DG. The effect of acute cellular glucoprivation induced by 2-DG (Sigma, St. Louis, MO) on extracellular concentrations of NE in the VMN, LHA, and PVN was evaluated. Dialysate samples were collected at 10-min intervals into microtubes containing 0.2 µl 0.1 N perchloric, and all samples were immediately frozen at the end of each sample period and maintained at 84°C until assayed. Blood samples (~100 µl) were collected at the midpoint of each 10-min period into 1-ml syringes and immediately transferred to chilled microtubes containing 10 µl heparin (500 U/ml). An equal volume of donor blood was infused after each sample. Baseline NE and plasma glucose concentrations were established during the 30 min preceding an infusion of saline or 2-DG (1.2 mmol/kg, dissolved in sterile saline at 1.2 mmol/0.5 ml).

Experiment 2: GABA or NE administration into the VMN or LHA. Baseline samples from the VMN (n = 8) and LHA (n = 7) were collected for 30 min before the perfusion media was changed to a KRB containing GABA (10 µM). Samples were collected at 10-min intervals for 30 min and analyzed for NE. A second trial evaluated the effect of exogenous NE on GABA concentrations in the VMN (n = 9), and LHA (n = 9) was evaluated. After baseline samples were collected for 30 min, a KRB containing NE (10 µM) was perfused through the microdialysis probes for 30 min, and samples were collected at 10-min intervals. Blood samples were collected at the midpoint of each 10-min sample collection period for measuring plasma glucose. In a third trial, NE (0.1 pg/0.5 µl artificial cerebrospinal fluid) was microinjected (0.1 µl/min) into the VMN (n = 5) after the baseline period via a combination microdialysis/injection probe. Dialysate samples were collected for 30 min after starting microinjection.

Experiment 3: Effect of yohimbine on GABA after 2-DG. The effects of the ₂-adrenoceptor antagonist yohimbine on GABA concentrations in the VMN (n = 11) and LHA (n = 11) after 2-DG were evaluated in this experiment. The effect of 2-DG was confirmed in a group of rats (n = 3) receiving 2-DG only. After baseline samples were collected for 30 min, a KRB with yohimbine (10 µM) was perfused through the microdialysis probes and samples were collected at 10-min intervals. After another 30-min period, 2-DG (1.2 mmol/kg, n = 7 per brain area) or saline (n = 4 per brain area) was administered through the jugular vein catheter, and samples were collected for 60 min. Blood samples were collected at the midpoint of each 10-min sample collection period for measuring plasma glucose.

Experiment 4: Effect of timolol on GABA after 2-DG. The effects of the -adrenoceptor antagonist timolol on GABA concentrations in the VMN (n = 11) and LHA (n = 11) after 2-DG was evaluated in this experiment. The effect of 2-DG was confirmed in a group of rats (n = 3) receiving 2-DG only. After baseline samples were collected for 30 min, a KRB containing timolol (10 µM) was perfused through the microdialysis probes, and samples were collected at 10-min intervals. After another 30-min period, 2-DG (1.2 mmol/kg, n = 7 per brain area) or saline (n = 4 per brain area) was administered through the jugular vein catheter, and samples were collected for 60 min. Blood samples were collected at the midpoint of each 10-min sample collection period for measuring plasma glucose.

Data analysis. Changes in neurotransmitter concentrations within a brain area and in plasma glucose concentration were analyzed by repeated-measures ANOVA. Differences in NE concentrations between saline and 2-DG were analyzed by Scheffé's multiple comparison tests and changes from baseline, within a treatment, were analyzed by paired t-test. The effect of GABA or NE on neurotransmitter concentrations within a brain area was determined by paired t-test. To determine the effect of adrenoceptor antagonists on extracellular GABA within a brain area, baseline concentrations with (30 to 0 min) and without (60 to 30 min) added antagonists were compared by paired t-test. The effect of the adrenoceptor antagonists on 2-DG-mediated changes in GABA concentrations was determined using repeated-measures ANOVA and Scheffé's post hoc comparison test. Results are presented as means ± SE.

Supplies. Ketamine, acepromazine, and butorphenol were obtained from Aveco (Fort Dodge, IA). Xylazine was obtained from Vedco (St. Joseph, MO). All other reagents were purchased from Sigma.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Experiment 1. Administered 2-DG was effective in inducing glucoprivation: in all animals, plasma glucose concentrations increased from 6.6 ± 0.3 to 12.1 ± 1.4 mM within 35 min of the rats receiving 2-DG. In the VMN (Fig. 1, top), NE concentrations increased [F(8,72) = 6.165, P < 0.001] in a bimodal fashion after 2-DG was administered. There was no effect of saline administration on extracellular NE concentrations. NE concentrations in dialysate increased to 251 ± 39% of baseline during the first 10-min sample period and 150 ± 17% of baseline 20-30 min after 2-DG administration. This bimodal pattern was apparent in all rats, although in two rats the second peak occurred during the 30- to 40-min sample period. Concentrations remained ~20% higher than baseline throughout the sample collection period. The induced glucoprivation also affected [F(8,72) = 7.92, P < 0.001] NE concentrations in the LHA (Fig. 1, middle). There was no effect of saline infusion; however, NE concentrations were reduced to 61 ± 4% of baseline during the 20- to 30-min sample and increased to 146 ± 18% of baseline during the last sample period. A decrease in NE concentrations was apparent in all animals, and the rebound was apparent in four of six animals. There was considerable variation and no consistent pattern [F(8,88) = 0.17, P = 0.99] in NE concentrations in the PVN (Fig. 1, bottom) after 2-DG. In five of seven rats evaluated, NE concentrations were at least 127% of baseline in at least one sample, occurring in the second sample period in two rats, the third sample period in one rat, and the fourth sample period in two rats. There was a decrease to near 55% of baseline in one sample period in four of seven rats, occurring during the second sample period in two rats. Baseline concentrations (pg/10 µl) of NE were 4.66 ± 1.12, 2.39 ± 0.17, and 7.80 ± 1.34 in the VMN, LHA, and PVN, respectively. Cannula placements are illustrated in Fig. 2.

View larger version (27K): [in this window] [in a new window]   Fig. 1.   Effect of saline () or 1.2 mmol/kg 2-deoxy-D-glucose (2-DG; ) on norepinephrine (NE) concentration in the ventromedial nucleus (VMN), lateral hypothalamic area (LHA), or paraventricular nucleus (PVN). Saline or 2-DG was administered (0.5 ml/kg) into the jugular vein at time 0. Data are presented as a percent of baseline. * Different (P < 0.05) from saline as determined by Scheffé's multiple comparison test and different from baseline by paired t-test. For each brain site, n = 5-7 rats per treatment .

View larger version (26K): [in this window] [in a new window]   Fig. 2.   Cannula placements in experiment 1 of rats receiving saline () or 1.2 mmol/kg 2-DG (). The ends of guide cannulas are represented by circles, and cannula and probe extensions into the VMN, LHA, or PVN are depicted by lines. AP, anteroposterior.

Experiment 2. In the VMN (Fig. 3), exogenous GABA had little effect [t(7) = 0.40, P = 0.88] on extracellular NE concentrations over the 30-min sample period. There was a 17 ± 6% decrease [t(7) = 2.79, P = 0.03] in NE concentrations in the first 10-min sample after adding GABA to KRB. In response to exogenous NE into the VMN (Fig. 3), there was an immediate increase in extracellular GABA concentration, which remained 66 ± 13% above baseline [t(8) = 5.21, P < 0.01] throughout the sample period. After microinjection of NE into the VMN, GABA concentrations were greater [t(4) = 3.62, P < 0.01] in the first two samples and near baseline levels during the third sample. In the LHA (Fig. 3), NE concentrations decreased from 4.13 ± 0.41 to 3.39 ± 0.29 pg/10 µl [t(6) = 3.00, P = 0.01] in response to exogenous GABA. GABA concentrations in the LHA were not affected by exogenous NE [t(8) = 1.23, P = 0.25]. Plasma glucose concentrations increased from 122 ± 8 to 133 ± 8 mg/dl [t(7) = 2.68, P = 0.04] when GABA was perfused through the VMN. Plasma glucose concentrations were not affected by administering GABA into the LHA or NE into either the VMN or LHA.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Effect of exogenous GABA on NE release (left, n = 8 or 9 per brain area) and exogenous NE on GABA release (right, n = 5 per brain area) in the VMN (open bars) or LHA (closed bars). Samples were collected for 30 min after GABA (10 µM) or NE (10 µM) were added to Krebs-Ringer buffer after establishing baseline concentrations of NE or GABA. ** Different from baseline (P < 0.01) as determined by paired t-test.

Experiment 3. The addition of yohimbine to KRB affected basal GABA concentrations in the VMN and LHA differently. The GABA response to 2-DG in both the VMN and LHA was confirmed in rats not receiving yohimbine. In the VMN (Fig. 4, top), yohimbine reduced [t(10) = 3,60, P < 0.01] GABA concentrations 18% (0.45 ± 0.10 to 0.37 ± 0.10 pmol/10 µl), whereas baseline GABA concentrations in the LHA (Fig. 4, bottom) were increased [t(10) = 6.21, P < 0.01] from 0.38 ± 0.07 to 0.57 ± 0.09 pmol/10 µl. Yohimbine did not prevent a 2-DG-induced increase [F(8,64) = 3.08, P < 0.01] in GABA concentrations in the VMN. GABA concentrations 20-30 min after 2-DG were increased to 248 ± 63% of baseline. This increase occurred in this time period in six of seven rats and during the 30- to 40-min period in one rat. There was no effect of 2-DG on GABA concentrations in the LHA, where GABA remained elevated as long as yohimbine was present in the KRB. There was no effect of yohimbine on baseline or 2-DG-induced increases in plasma glucose when supplied to either brain area.

View larger version (38K): [in this window] [in a new window]   Fig. 4.   Effect of yohimbine on extracellular GABA concentration in the VMN (top) and LHA (bottom). Yohimbine (10 µM) was added to Krebs-Ringer buffer after baseline concentrations were established and 30 min before 2-DG (1.2 mmol/kg iv) or equal volume of sterile 0.9% saline was administered at time 0. The effect of 2-DG was confirmed in a group of rats receiving 2-DG only. There was a significant effect of 2-DG in the VMN [F(8,64) = 3.08, P < 0.01] but not in the LHA [F(8,64) = 1.78, P = 0.10]. ** Different (P < 0.01) from saline as determined by Scheffé's multiple comparison test and different from baseline by paired t-test. # Different (P < 0.05) from baseline by paired t-test. For each brain site, there were n = 4 rats treated with saline and n = 7 or 8 rats treated with 2-DG.

Experiment 4. The addition of timolol to KRB produced similar results as yohimbine: basal GABA concentrations were reduced [t(9) = 1.55, P = 0.15] in the VMN (Fig. 5, top) and were 42% higher [t(10) = 4.25, P < 0.01] in the LHA (Fig. 5, bottom). The GABA response to 2-DG in both the VMN and LHA was confirmed in rats not receiving timolol. In the presence of timolol, 2-DG invoked an increase [F(8,64) = 3.26, P < 0.01] in GABA concentrations in the VMN. GABA concentrations increased to 142 ± 6% of baseline during the first 10-min sample period after 2-DG was administered. The increase during this time period occurred in all seven rats. There was no effect of 2-DG on GABA in the LHA nor did timolol affect baseline or 2-DG-induced increases in plasma glucose when supplied to either brain area.

View larger version (37K): [in this window] [in a new window]   Fig. 5.   Effect of timolol on extracellular GABA concentration in the VMN (top) and LHA (bottom). Yohimbine (10 µM) was added to Krebs-Ringer buffer after baseline concentrations were established and 30 min before 2-DG (1.2 mmol/kg iv) or equal volume of sterile 0.9% saline was administered at time 0. The effect of 2-DG was confirmed in a group of rats receiving 2-DG only. There was a significant effect of 2-DG in the VMN [F(8,64) = 3.26, P < 0.01] but not in the LHA [F(8,64) = 1.49, P = 0.10]. ** Different (P < 0.01) from saline as determined by Scheffé's multiple comparison test and different from baseline by paired t-test. # Different (P < 0.05) from baseline by paired t-test. For each brain site, there were n = 4 rats treated with saline and n = 7 rats treated with 2-DG).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Noradrenergic activity in the ventromedial hypothalamus is increased in response to an acute 2-DG-induced glucoprivic episode. The present results are consistent with earlier reports of 2-DG-induced increases in extracellular NE concentration in the medial hypothalamus (17, 18, 28); however, the timing and pattern of changes differs among these reports. The bimodal pattern of increased NE we measured with a 10-min sampling interval differed from the broad peak of NE between 30 and 90 min reported using 30-min samples (28). An immediate increase (e.g., first 5-min sample) in extracellular NE in the medial hypothalamus was also apparent using push-pull perfusion techniques after 2-DG was administered peripherally (17) or directly into the medial hypothalamus (18). The increase in extracellular NE likely reflects increased noradrenergic activity in the ventromedial hypothalamus. NE turnover in the ventromedial hypothalamus was increased in response to a glucoprivic episode (30, 31).

The bimodal increase in extracellular NE in the VMN mirrored a bimodal increase in extracellular GABA in the VMN after 2-DG (2). The increased noradrenergic activity may mediate the increase in extracellular GABA with each of the peaks in NE influencing GABA release by a different adrenoceptor subtype. Because the initial peak in GABA concentration was absent when yohimbine was administered into the VMN, the immediate increase in GABA after 2-DG is dependent on activation of ₂-adrenoceptors. The second rise in GABA activity, 20-30 min after 2-DG, was not affected by yohimbine but was absent in the presence of timolol. Thus it appears the second GABA peak is mediated by -adrenoceptors. These results support the interaction of NE and GABA systems in the medial hypothalamus in the maintenance of plasma glucose homeostasis.

The noradrenergic response in other hypothalamic nuclei to 2-DG is less clear. Decreased extracellular NE was measured 30 min after 2-DG in the LHA in this and another study (28). Conversely, NE concentrations were reported to be higher in the LHA after intravenous (13) or direct infusion of 2-DG into the LHA (20, 24). In general, NE concentrations in the medial hypothalamus were increased during the first 5 min after peripherally administered 2-DG was administered (17) or 2-DG was applied directly to the hypothalamus (18). These reports combined data from the VMN, DMN, and PVN and included no discussion of a difference in responses among hypothalamic sites. In a push-pull perfusion study, Peinado and Myers (24) measured a rapid increase in NE release in the PVN when 2-DG was administered directly at the sample site. In the present study, there was no consistent change in NE release in the PVN after 2-DG nor did Shimizu and Bray (28) observe an effect of 2-DG on NE release in the DMN.

The bulk of the present data supports NE mediation of GABAergic activity in the VMN; however, a slight transient decrease in NE release after GABA administration into the VMN was observed. This in vivo data support evidence of NE mediating GABA release in vitro. In synaptosomal preparations from hypothalamic tissue, NE stimulated GABA release in a concentration-dependent manner (16). The effect of NE on GABA release was mimicked by clonidine, an ₂-adrenoceptor agonist, and blocked by yohimbine, an ₂-adrenoreceptor antagonist (16, 25). The ₂-adrenoceptors appeared to be on the GABA nerve terminals (16). Electrical stimulation of the A1 nuclear group increased extracellular NE and GABA in the preoptic area (POA) of the hypothalamus (9). The increase in GABA was likely to have been influenced by the increased NE as the -antagonist phenoxybenzamine blocked NE-induced increases in GABA in POA slices (8). An effect of GABA on NE release in the hypothalamus has also been reported, and, in the present study, exogenous NE reduced GABA activity in the lateral hypothalamus. Whether the effect was via GABA_A or GABA_B receptors was not evaluated, and both receptor types have been reported to influence NE release. Navarro et al. (21) reported that GABA reduced glutamate-stimulated NE release from superfused pieces of the medial basal hypothalamus via GABA_B receptors. In hypothalamic slices, NE release was increased in the presence of the GABA_A receptor agonist muscimol (5, 6). Because muscimol did not affect NE release from synaptosomes, the authors suggested that GABA's effect was through interneurons (5).

The interaction among GABAergic and noradrenergic systems in the LHA appears contradictory. Exogenous GABA reduced NE concentrations in the LHA, and there was a decrease in NE in the LHA 20-30 min after 2-DG. However, because GABA concentrations were not increased in the LHA after 2-DG (2), the decrease in NE concentration in the LHA was unlikely the result of endogenous GABA. Furthermore, extracellular GABA was increased in the LHA in the presence of either adrenoceptor antagonist, and NE in the LHA was reduced after 2-DG. However, GABA concentrations were not increased in the LHA after 2-DG (2), so it is also unlikely that changes in NE were mediating the changes in GABA in vivo. The clear difference in basal GABA concentrations in the VMN and LHA when either adrenoceptor antagonist was present at the same concentration supports the GABA neurons in these two areas being different. The functional significance of the interaction of NE and GABA in the LHA is unclear. Blood glucose levels were not affected by infusion of either antagonist into the LHA, although perfusions were unilateral.

It is likely that the increase in noradrenergic activity in the VMN after 2-DG is involved in initiating compensatory responses to glucoprivation. Microinjection of NE into the medial hypothalamus stimulates food intake (19), increases the firing rate of sympathetic efferents to brown adipose tissue (26), and increases circulating glucose and glucose-mobilizing hormones (4, 32). The effect of NE in the medial hypothalamus to increase plasma glucose is via ₂-adrenoceptors (4), which appear to be upregulated by 2-DG (1). Administering either phentolamine (19) or yohimbine (29) blocked the hyperglycemic response to 2-DG. After 2-DG administration, animals preferentially consume carbohydrate (11), a response mediated by ₂-adrenoceptors in the medial hypothalamus (14). Blocking the increase in hypothalamic NE activity via pentobarbital sodium (30) or -methylparatyrosine (19) prevented the feeding response to glucoprivation. Injection of NE receptor antagonists into the VMN also blocked the feeding response to 2-DG (19, 30). Confining glucoprivation to the medial hypothalamus by direct application of 2-DG was enough to elicit increased NE in VMN (18) and compensatory physiological responses (e.g., increased plasma glucagon and epinephrine) (3). In the present study, adding yohimbine or timolol to the dialysate had little effect on the plasma glucose response to 2-DG. The lack of an effect by either antagonist was likely the result of the unilateral application. Previous studies documenting effects on plasma glucose of these antagonists were administered bilaterally or into a lateral ventricle.

The functional significance of the bimodal pattern in neurotransmitter activity in the VMN and differential affects of adrenoceptor type is unclear. The two peaks may be involved in different aspects of glucose mobilization. Scheurink et al. (27) noted differences in the role of VMN - and -adrenoceptors on sympathetic activity during exercise. An exercise-induced increase in plasma epinephrine was reduced by -blockade in the VMN, whereas the increase in plasma NE concentration during exercise was reduced when timolol was administered into the VMN. VMN -adrenoceptors influence plasma glucose via sympathetic-mediated effects on hepatic phosphorylase activity and glycogenolysis (15). In our previous study (2), the timing of the second peak in GABA concentration 20-30 min after 2-DG was associated with feeding and was not blocked by yohimbine, although the feeding response to 2-DG has been blocked by -, but not -, adrenoceptor antagonists (19). In the present study, it was the first peak in GABA concentrations that was mediated by ₂-adrenoceptors. A functional GABAergic system in the VMN is necessary for feeding after 2-DG: application of the GABA receptor antagonist bicuculline into the VMN blocked 2-DG-induced feeding (10). Coadministering phentolamine, an -adrenergic antagonist, with muscimol did not block the feeding response to muscimol; however, the increased food intake after NE injection was blocked by coinjecting bicuculline, a GABA_A receptor antagonist (7).

Perspectives