Perifornical hypothalamic orexin and serotonin modulate the counterregulatory response to hypoglycemic and glucoprivic stimuli

Abstract

Previous reports suggested an important role for serotonin (5-hydroxytryptamine [5-HT]) in enhancing the counterregulatory response (CRR) to hypoglycemia. To elucidate the sites of action mediating this effect, we initially found that insulin-induced hypoglycemia stimulates 5-HT release in widespread forebrain regions, including the perifornical hypothalamus (PFH; 30%), ventromedial hypothalamus (34%), paraventricular hypothalamus (34%), paraventricular thalamic nucleus (64%), and cerebral cortex (63%). Of these, we focused on the PFH because of its known modulation of diverse neurohumoral and behavioral responses. In awake, behaving rats, bilateral PFH glucoprivation with 5-thioglucose stimulated adrenal medullary epinephrine (Epi) release (3,153%) and feeding (400%), while clamping PFH glucose at postprandial brain levels blunted the Epi response to hypoglycemia by 30%. The PFH contained both glucose-excited (GE) and glucose-inhibited (GI) neurons; GE neurons were primarily excited, while GI neurons were equally excited or inhibited by 5-HT at hypoglycemic glucose levels in vitro. Also, 5-HT stimulated lactate production by cultured hypothalamic astrocytes. Depleting PFH 5-HT blunted the Epi (but not feeding) response to focal PFH (69%) and systemic glucoprivation (39%), while increasing PFH 5-HT levels amplified the Epi response to hypoglycemia by 32%. Finally, the orexin 1 receptor antagonist SB334867A attenuated both the Epi (65%) and feeding (47%) responses to focal PFH glucoprivation. Thus we have identified the PFH as a glucoregulatory region where both 5-HT and orexin modulate the CRR and feeding responses to glucoprivation.

Figure 1

5-HT turnover in selected brain regions after IIH. Rats (n = 5–6/group) were given bolus insulin (4.5 units/kg s.c.) or vehicle (saline) injections, and 5-HT turnover (ratio of 5-HIAA to 5-HT) was assessed in brain micropunches after 2 h of hypoglycemia. Bars are mean ± SEM. *P = 0.05 or less between hypoglycemia and saline groups by one-way ANOVA with Tukey post hoc correction. Ctx, cerebral cortex; DMN, dorsomedial hypothalamic nucleus; dRa, dorsal raphe nucleus; vHip, ventral hippocampus.

Figure 2

Effects of local PFH glucoprivation on glucoregulatory hormone release. Rats (n = 4–8/group) were infused bilaterally with 5-TG (60 μg in 0.5 μL) or vehicle (saline) in the PFH. Results are shown for blood glucose (A), epinephrine (B), norepinephrine (C), and glucagon (D). Data points are mean ± SEM. *P = 0.05 or less between 5-TG and saline groups at each time point by t test after responses over 2 h were significant by one-way repeated-measures ANOVA. Glucagon levels in saline controls (2D) were below assay detection limits.

Figure 3

Effect of clamping PFH extracellular glucose at euglycemic levels on the CRR to IIH. Rats (n = 7–8/group) were given bolus insulin (4.5 units/kg i.v.) and reverse dialyzed with 25 mmol/L glucose or ECF. Results are shown for blood epinephrine (A), norepinephrine (B), glucagon (C), and glucose (D). Data points are mean ± SEM. *P = 0.05 or less between 25 mmol/L glucose and ECF groups at each time point by t test after responses over 2 h were significant by one-way repeated-measures ANOVA.

Figure 4

Effect of PFH ablation of 5-HT signaling on glucoregulatory hormone release after systemic and local PFH glucoprivation. Rats (n = 6–10/group), previously treated with 5,7-DHT (5 μg in 0.5 μL infused over 5 min, in combination with 25 mg/kg desmethylimipramine s.c.) or vehicle (0.1% ascorbic acid), were given systemic 2-DG (200 mg/kg) (A). In a separate group, rats (n = 5/group) with prior 5,7-DHT lesions or sham operations, were infused with 5-TG (60 μg in 0.5 μL) into the PFH (B–D). Data points are mean ± SEM. *P = 0.05 or less between 5,7-DHT and sham-operated groups at each time point by t test after responses over 2 h were significant by one-way repeated-measures ANOVA.

Figure 5

Effect of PFH sertraline reverse dialysis on the CRR to IIH. Rats (n = 5–6/group) were reverse dialyzed with sertraline (10 μmol/2 h), sertraline + 5-HT (10 μmol/2 h + 10 nmol/2 h, respectively), or vehicle (3% DMSO, 0.1% ascorbic acid) in combination with systemic IIH (4.5 units/kg bolus insulin i.v.). Results are shown for blood epinephrine (A), norepinephrine (B), glucagon (C), and glucose (D). Data points are mean ± SEM. Curves marked by different letters are significantly different by one-way repeated-measures ANOVA.

Figure 6

Effect of 5-HT on PFH glucosensing neurons in vitro. PFH neurons were classified as GE or GI and, at glucose levels comparable to those seen during hypoglycemia (0.5 mmol/L), were assessed for their responses to 50 pmol/L 5-HT. Bars are mean ± SEM. SerE, 5-HT excited; SerI, 5-HT inhibited; SerN, 5-HT nonresponsive.

Figure 7

Effect of systemically administered brain-penetrant orexin 1 receptor antagonist SB334867A on PFH glucoprivation-induced glucoregulatory hormone release. Rats (n = 5–6/group) received bilateral PFH 5-TG infusions (60 μg in 0.5 μL) in combination with systemic administration of SB334867A (20 mg/kg i.v.) or vehicle (100% DMSO). Results are shown for blood glucose (A), epinephrine (B), norepinephrine (C), and glucagon (D). Data points are mean ± SEM. *P = 0.05 or less between SB334867A and DMSO groups at each time point by t test after responses over 2 h were significant by one-way repeated-measures ANOVA.