Blockade of GABA_A Receptors in the Ventromedial Hypothalamus Further Stimulates Glucagon and Sympathoadrenal but Not the Hypothalamo-Pituitary-Adrenal Response to Hypoglycemia

Male Sprague-Dawley rats (Charles River, Wilmington, MA) weighing 300–350 g were individually housed in the Yale Animal Resource Center in temperature- (22–23°C) and humidity-controlled rooms. The animals were fed rat chow (Agway Prolab 3000; Syracuse, NY) and water ad libitum and were acclimatized to a 12-h light-dark cycle (lights on between 0700 and 1900) for a period of 1 week before experimental manipulation. Principles of laboratory animal care were followed, and experimental protocols were approved by the Yale University Institutional Animal Care and Use Committee.

Three groups of rats were used: control rats administered artificial extracellular fluid (aECF) (n = 8), rats administered the GABA_A receptor antagonist bicuculline methiodide (n = 8), and rats administered the GABA_A receptor agonist muscimol (n = 10). On day 0, vascular catheters were placed into the left carotid artery and right jugular vein, as described previously (38). In addition, bilateral stainless steel microinjection guide cannulas were stereotaxically positioned 1 mm dorsal to the VMH according to the coordinates of Paxinos and Watson (from bregma: −2.6 mm anterior posterior, ±3.8 mm medial lateral, and −8.3 mm dorsal ventral, at an angle of 20°) (39). The animals were then allowed 7–10 days to recover from surgery and to acclimatize to being handled. Those animals that did not recover to their presurgery body weights were excluded from the study. Following an overnight fast, the experiments were carried out in conscious and unrestrained animals. Catheters were extended outside of the cage to minimize investigator interaction and were connected to infusion pumps. The animals were allowed ∼2–2.5 h to recover from handling before basal hormone samples were collected (between 0930 and 1000), just before the start of the study. Following the collection of basal blood samples, microinjection needles were inserted down to the level of the VMH, and 0.1 μl aECF (135 mmol/l NaCl, 3 mmol/l KCl, 1 mmol/l MgCl₂, 1.2 mmol/l CaCl_2, and 2 mmol/l Na₂HPO₄) and 0.75 mmol/l glucose (40,41), bicuculline methiodide in aECF (12.5 pmol/side), or muscimol in aECF (1 nmol/side) was delivered over the course of 1 min using a CMA/102 pump (CMA Microdialysis, N. Chelmsford, MA). The dose of bicuculline methiodide was achieved by gradually titrating down a concentration of the compound that caused convulsions until reaching a level that was just below the threshold where convulsions were consistently observed. With muscimol, the lowest dose that produced a consistent effect was used. Pilot studies using India ink verified that a volume of 0.1 μl delivered over 1 min was adequately localized to the VMH region. The rats then underwent a hyperinsulinemic-hypoglycemic glucose clamp. A constant insulin (50 mU · kg⁻¹ · min⁻¹ regular human insulin; Eli Lilly, Indianapolis, IN) and variable dextrose (20%; Abbott, Chicago, IL) infusion through the jugular vein catheter was used to maintain plasma glucose levels at 2.5 ± 0.3 mmol/l for 90 min. The concentration of glucose was determined every 5 min from a sample of plasma (10 μl) using a Beckman glucose analyzer II (Beckman, Palo Alto, CA). Arterial blood samples were obtained from the carotid catheter at 0, 30, 60, 90, and 120 min following the start of the insulin infusion. For glucagon and insulin measurements, 250 μl whole blood was collected in chilled tubes containing EDTA (J.T. Baker, Phillipsburg, NJ) and Trasylol (Bayer, Tarrytown, NY). Blood samples for catecholamines (250 μl) were collected in chilled tubes containing 4 μl reduced glutathione/EGTA (Amersham, Arlington Heights, IL). For corticosterone measurements, 250 μl whole blood was collected into an eppendorf tube and serum separated from the erythrocytes. The plasma and serum were aliquoted into storage tubes and stored at −20°C (or −80°C for catecholamine determination). Erythrocytes, after removal of plasma, were resuspended in an equivalent volume of sterile artificial plasma (115 mmol/l NaCl, 5.9 mmol/l KCl, 1.2 mmol/l MgCl₂, 1.2 mmol/l NaH₂PO₄, 1.2 mmol/l Na₂SO₄, 2.5 mmol/l CaCl₂, 25 mmol/l NaHCO₃, and 4% BSA, pH 7.45) and reinfused after each blood sampling to prevent volume depletion and anemia.

To test the specificity of our manipulations on counterregulatory responses to hypoglycemia, we performed a set of hyperinsulinemic-euglycemic (6.4 ± 0.6 mmol/l) clamps in two groups of animals: control rats administered aECF (n = 5) and rats administered the GABA_A receptor agonist muscimol (n = 6). As described above, these compounds were microinjected into the VMH just before the start of constant insulin (50 mU · kg⁻¹ · min⁻¹) and variable dextrose (50%; Abbott) infusions. The same blood sampling protocol as described above was followed.

At the end of the experiments, the rats were killed with an overdose of sodium pentobarbital (Sleepaway; Fort Dodge Animal Health, Fort Dodge, IO). Brains were removed, frozen, and stored at −80°C until analysis. Accuracy of microinjection needle placements were verified histologically in coronally sectioned brain slices at the end of the study. A schematic of typical needle placements is shown in Fig. 1. Only data collected from those animals with probes in the appropriate target area were included for statistical analysis.

Plasma insulin (Linco Research, St. Charles, MO), corticosterone (Diagnostic Products, Los Angeles, CA), and glucagon (Linco Research) concentrations were determined using commercially available radioimmunoassay kits. Plasma epinephrine and norepinephrine concentrations were analyzed by high-performance liquid chromatography using electrochemical detection (ESA, Acton, MA).

Hormonal data are presented as means ± SE. Statistical analysis was performed by one- or two-way ANOVA for independent or repeated measures, as appropriate, followed by post hoc analysis using Statistica 6.0 (StatSoft, Tulsa, OK) for personal computers. P < 0.05 was set as the criterion for statistical significance.

Plasma hormone concentrations at baseline were not significantly different among experimental treatment groups (Table 1). During the insulin infusion, plasma glucose was reduced to similar levels in all three treatment groups (Fig. 2), and average plasma insulin concentrations during the clamping period were also similar between treatment groups (Fig. 3A). Nevertheless, the exogenous glucose infusion rates required to achieve this level of glycemia were significantly altered by local changes in GABAergic tone within the VMH region. Significantly less (P < 0.05) glucose was needed in animals treated with the GABA_A receptor antagonist bicuculline methiodide, and significantly more (P < 0.05) was needed in animals receiving the GABA_A receptor agonist muscimol, as compared with the control group (Fig. 3B).

VMH delivery of the GABA_A receptor antagonist increased peak plasma glucagon responses by 62% (P < 0.04), whereas VMH delivery of the GABA_A receptor agonist reduced peak glucagon responses by 44% (P < 0.05) compared with vehicle controls (Fig. 4). A similar pattern was observed for epinephrine. Peak epinephrine responses were significantly higher (P < 0.001) with bicuculline methiodide and lower with muscimol (P < 0.05) injection compared with controls (Fig. 5). On the other hand, plasma norepinephrine concentrations were not significantly affected.

Peak plasma corticosterone concentrations did not differ significantly between the three treatment groups during the hypoglycemic clamp (Fig. 6).

Baseline plasma hormone concentrations were not significantly different between the two experimental groups (Table 2). Plasma glucose levels in the two treatment groups were maintained at similar levels during the euglycemic clamp (Fig. 7). Neither average insulin concentrations (control 4,071 ± 129 and muscimol 4,168 ± 388 μU/ml) nor glucose infusion rates (controls 33 ± 1 and muscimol 32 ± 2 mg · kg⁻¹ · min⁻¹) differed between the two groups during the euglycemic clamping period.

Local delivery of the GABA_A receptor agonist muscimol into the VMH did not significantly affect any of the counterregulatory hormone responses during the hyperinsulinemic-euglycemic clamp (Figs. 8–10).

The mechanisms underlying blood glucose sensing in the brain remain elusive. Recent studies indicate that several regions within the brain may serve as sensors to detect changes in glucose levels, including the paraventricular nucleus, the arcuate nucleus, and the VMH (11–13,19,42). The latter two regions contain neurons that not only possess much of the glucose-sensing machinery (i.e., glucokinase, glucose transporters, and K_ATP channels) but also adjust their firing rates in accordance with changes in ambient glucose levels, either increasing (glucose-excited neurons) or decreasing (glucose-inhibited neurons) rates in response to high glucose levels (18,42,43). In studies conducted by Kang et al. (25) to characterize VMH neurons, it was noted that 56% of glucose-excited and 36% of glucose-inhibited neurons expressed GAD, the rate-limiting enzyme in GABA synthesis. This observation suggests that GABA could be a potential candidate neurotransmitter modulating brain glucose sensing and, in turn, glucose homeostasis. Early studies showed that hepatic glucose production in the basal state can be enhanced with central GABA antagonism (44), providing some evidence for the role of GABA in regulating glucose homeostasis. Whether GABA plays a predominant role in modulating glucose counterregulatory responses to hypoglycemia is still unclear. Measurements of GABA in the brain during hypoglycemia show varying results that may be indicative of regionally specific regulation. For example, perfusion of 2-deoxyglucose or systemic insulin-induced hypoglycemia produced a 45 and 49% decrease, respectively, in microdialysate GABA concentration in the substantia nigra (31). On the other hand, insulin-induced hypoglycemia has been reported to cause a 47% increase in hippocampal GABA levels (45). In the VMH, a biphasic increase in extracellular GABA levels was reported following 2-deoxyglucose administration and insulin-induced hypoglycemia (32,33). Thus, although it is likely that hypoglycemia elicits a multifaceted change in GABAergic tone within the central nervous system, the reason for these differing responses remains to be established.

To clarify the role of GABA in hypoglycemic counterregulation, we microinjected a GABA receptor agonist and antagonist specifically into the VMH to establish whether changes in GABAergic inhibitory tone within this region can affect glucagon and sympathoadrenal responses to hypoglycemia. Our data demonstrate that local VMH administration of the GABA receptor antagonist amplified glucagon and epinephrine responses during hypoglycemia, whereas the agonist suppressed them. Norepinephrine responses, on the other hand, were not affected by either the antagonist or agonist. These results are consistent with previous studies conducted under basal conditions that revealed a rise in glucagon and epinephrine levels when bicuculline methiodide was injected into the third ventricle (29). Although this study did not localize the brain region responsible for changes in glucoregulatory hormones, our data suggest that the GABAergic system within the VMH is primarily responsible for these effects. However, we cannot establish with certainty how the GABAergic system is integrated into the entire glucose-sensing system or, more specifically, how its effects on glucagon and epinephrine secretion are mediated. It is notable that Nonogaki et al. (29) suggested that the rise in plasma glucagon levels following bicuculline methiodide administration may have been mediated indirectly by a corresponding increase in epinephrine secretion, since no significant changes in glucagon secretion were observed when the experiment was repeated in adrenalectomized rats. In contrast to these observations, DeRosa and Cryer (46) reported that the glucagon response to hypoglycemia is still intact in bilaterally adrenalectomized patients, suggesting that epinephrine may not be an absolute requirement for glucagon release in humans. Although species differences could account for the conflicting results in the two studies, we hypothesize that GABA released from glucose-sensing neurons within the VMH may act to modulate sympathetic nervous system output to pancreatic α-cells and the adrenal medulla and thereby play a role in determining the magnitude of these counterregulatory responses to hypoglycemia. Whereas our data provides evidence for the role of GABA in the VMH in modulating glucagon and sympathoadrenal responses to hypoglycemia, it does not necessarily exclude other brain regions or even other substrates or neurotransmitters from playing a contributory role to glucose counterregulation.

There is currently little information available regarding the specific role of glutamate, an excitatory neurotransmitter, in regulating counterregulatory responses to hypoglycemia. Studies have reported that stimulation of glutamate receptors in the VMH can increase peripheral glucose disposal (47) and activate sympathetic nervous activity in intrascapular brown adipose tissue (48). These studies suggest that glutamate could potentially have a role in the regulation of glucose and energy homeostasis. In fact, GABAergic neurons in the VMH not only synapse on other neurons immunoreactive for GABA, but they also synapse on other as yet unidentified neurons. It is possible that glutamatergic tone may also contribute to modulation of counterregulatory responses. Although the study of glutamate is beyond the scope of the current study, its contribution to glucose counterregulation warrants further investigation.

While plasma glucose levels fell at similar rates in the three different treatment groups in the current study, local VMH delivery of a GABA receptor antagonist lowered and a GABA agonist raised glucose infusion rates. These observations are consistent with data showing that intracerebroventricular bicuculline methiodide administration altered baseline glucose homeostasis, producing hyperglycemia in a dose-dependent manner (28,29) and stimulating endogenous glucose production (44). Our data further extend the findings of these studies by examining the GABAergic system under conditions of hypoglycemia and, more importantly, implicate the VMH as the brain center responsible for these effects. We postulate that the observed changes in glucose infusion rates with GABA agonism and antagonism are likely due in part to the effects of glucagon and epinephrine on peripheral glucose production and utilization.

Neither VMH pretreatment with the GABA_A receptor agonist or antagonist altered pituitary-adrenal responses to hypoglycemia. In contrast, Giordano et al. (34) reported that activation of GABA receptors with the benzodiazepine Alprazolam can suppress ACTH and cortisol responses to hypoglycemia in humans when taken systemically. Taken systemically, the benzodiazepine can potentially activate GABAergic receptors in other sites within the central nervous system. More specifically, if activated, GABA receptors in the hypothalamic paraventricular nucleus, which is believed to be the locus for initiation of the stress response, can suppress hypothalmo-pituitary-adrenal (HPA) activity (35). Our findings suggest that GABAergic tone within the VMH may not play a crucial role in regulating the HPA response to hypoglycemia. In agreement with this, Evans et al. (49) demonstrated that lidocaine-induced inactivation of neurons within the dorsomedial hypothalamus was sufficient to inhibit HPA responses, but not the glucagon or epinephrine responses, to hypoglycemia. These data are consistent with the idea that activation of the HPA axis in response to hypoglycemia may require signals that emanate from neurons residing outside of the VMH and possibly from the dorsomedial nucleus and/or the paraventricular nucleus itself. In accordance with this hypothesis, we noted that manipulation of GABAergic tone within the hypothalamic paraventricular nucleus specifically altered adrenocortical responses to hypoglycemia and not to those of glucagon or epinephrine (O.C., unpublished observations). This may indicate that activation of the HPA axis and the sympathoadrenal system in response to hypoglycemia may be regulated by different regions within the brain, consistent with the idea of a redundant failsafe mechanism.

While feeding behavior was not examined in the current study, it is well known that insulin-induced hypoglycemia and 2-deoxyglucose elicit very clear hyperphagic responses. Although GABA in the VMH does appear to have a role in stimulating feeding behavior (50), GABAergic neurons in the VMH form many different synaptic connections that do not all originate from glucose-sensitive neurons. It is unclear whether the neurons involved in stimulation of counterregulatory responses are the same neurons that regulate feeding. It is possible that feeding behavior and modulation of counterregulatory responses are regulated by different subpopulations of GABAergic neurons within the VMH.

Interestingly, our euglycemia studies revealed that increasing GABAergic tone in the VMH under conditions of hyperinsulinemic-euglycemia does not significantly alter the secretion of counterregulatory hormones. This data suggests that counterregulatory responses may be under tonic inhibition under conditions of euglycemia. We hypothesize that changes in tonic inhibition during hypoglycemia may be one means by which the brain modulates the magnitude of counterregulatory responses when glucose levels fall.

Taken together, our data suggest that the GABAergic system within the VMH plays an important role in modulating the magnitude of glucagon and epinephrine responses to insulin-induced hypoglycemia. It is foreseeable that in situations where this system is disrupted or is ineffective, problems can potentially arise in one’s ability to mount an adequate counterregulatory response. Hence, this study has important implications in conditions such as long-standing diabetes, where both glucagon and catecholamine responses to hypoglycemia are either absent or significantly compromised. It is conceivable that an increase in GABAergic tone in the VMH, such as in the case of experimental diabetes (51), and/or an inability to relieve inhibitory tone from this hypothalamic region may contribute to impairment of the primary mechanisms that defend the body from hypoglycemia. As hypoglycemia remains the limiting factor in the proper glycemic management of patients with diabetes (2), understanding the mechanisms underlying impaired counterregulation may lead to the development of new treatment strategies that will minimize or prevent exposure to hypoglycemia. We speculate that minimizing GABAergic inhibitory tone in the VMH may be one means to improve counterregulatory responses in patients with diabetes.

This work was generously supported by research grants from the National Institutes of Health (DK20495) and the Juvenile Diabetes Research Foundation Center for the Study of Hypoglycemia. O.C. is a recipient of the Canadian Institutes of Health Research Post-Doctoral Research Fellowship.

The authors are grateful to Ralph Jacob, Aida Groszmann, and Andrea Belous for excellent technical support and assistance.