Ventromedial hypothalamic nitric oxide production is necessary for hypoglycemia detection and counterregulation

Abstract

Objective: The response of ventromedial hypothalamic (VMH) glucose-inhibited neurons to decreased glucose is impaired under conditions where the counterregulatory response (CRR) to hypoglycemia is impaired (e.g., recurrent hypoglycemia). This suggests a role for glucose-inhibited neurons in the CRR. We recently showed that decreased glucose increases nitric oxide (NO) production in cultured VMH glucose-inhibited neurons. These in vitro data led us to hypothesize that NO release from VMH glucose-inhibited neurons is critical for the CRR.

Research design and methods: The CRR was evaluated in rats and mice in response to acute insulin-induced hypoglycemia and hypoglycemic clamps after modulation of brain NO signaling. The glucose sensitivity of ventromedial nucleus glucose-inhibited neurons was also assessed.

Results: Hypoglycemia increased hypothalamic constitutive NO synthase (NOS) activity and neuronal NOS (nNOS) but not endothelial NOS (eNOS) phosphorylation in rats. Intracerebroventricular and VMH injection of the nonselective NOS inhibitor N(G)-monomethyl-l-arginine (l-NMMA) slowed the recovery to euglycemia after hypoglycemia. VMH l-NMMA injection also increased the glucose infusion rate (GIR) and decreased epinephrine secretion during hyperinsulinemic/hypoglycemic clamp in rats. The GIR required to maintain the hypoglycemic plateau was higher in nNOS knockout than wild-type or eNOS knockout mice. Finally, VMH glucose-inhibited neurons were virtually absent in nNOS knockout mice.

Conclusions: We conclude that VMH NO production is necessary for glucose sensing in glucose-inhibited neurons and full generation of the CRR to hypoglycemia. These data suggest that potentiating NO signaling may improve the defective CRR resulting from recurrent hypoglycemia in patients using intensive insulin therapy.

FIG. 1.

Decreased glucose increases VMH NO release. A: Representative trace of ex vivo amperometric measurements of NO release from mouse hypothalamus in response to an extracellular glucose decrease from 2.5 to 0.1 mmol/l. B: Mean frequency and (C) mean amplitude of NO release calculated during the last 10-min recording for each glucose level (n = 4). *P < 0.05 vs. 2.5 mmol/l glucose (one-way ANOVA).

FIG. 2.

Hypoglycemia increases ventral hypothalamic nNOS activity. A: Ventral hypothalamic constitutive (e/nNOS) or inducible (iNOS) NOS activity from rats injected subcutaneously with saline (control, n = 6) or insulin (2 units/kg; n = 6) 60 min after injection. B: Representative Western blot (left panel) of ventral hypothalamic total nNOS, phosphorylated nNOS (P-nNOS), total eNOS, and P-eNOS from control or insulin-treated rats injected subcutaneously with saline (n = 5) or insulin (n = 5) 60 min after injection. The right panel shows the quantification of the ratio between P-nNOS or P-eNOS and total nNOS or eNOS, respectively. Data are means ± SEM and represented as percentage of saline where the control group was considered to be 100%. *P < 0.05 vs. control (unpaired t test).

FIG. 3.

VMH NO signaling is necessary for recovery to euglycemia after insulin-induced hypoglycemia. Blood glucose levels in response to insulin-induced hypoglycemia (1 unit/kg, i.v.) in rats receiving (A) ICV perfusion of aCSF (controls; n = 14) or l-NMMA (50 mmol/l; n = 14), or (B) unilateral VMH injection of aCSF (n = 7), l-NMMA (50 mmol/l, n = 7), or ODQ (0.1 mmol/l, n = 5). *, #P < 0.05 vs. control (two-way ANOVA).

FIG. 4.

VMH NOS inhibition impairs the CRR to hypoglycemia. Blood glucose level (A); GIR (B); plasma epinephrine (C), and glucagon levels (D) during hyperinsulinemic/hypoglycemic clamp (1.2 units · kg⁻¹ · h⁻¹) of animals injected bilaterally in the ventromedial hypothalamus with aCSF (controls; n = 8) or l-NMMA (50 mmol/l; n = 6). *P < 0.05 vs. controls (two-way ANOVA).

FIG. 5.

nNOS is necessary for full initiation of the CRR. Blood glucose concentration (A), glucose infusion rate (B), and plasma epinephrine taken at the end of the clamp (C) of wild-type (WT) (n = 14), eNOS (n = 6), and nNOS (n = 7) knockout mice during hyperinsulinemic/hypoglycemic clamp (1.2 units · kg⁻¹ · h⁻¹). *P < 0.05 vs. wild type (two-way ANOVA).

FIG. 6.

nNOS is necessary for glucose sensing by VMN glucose-inhibited neurons. Representative whole-cell current-clamp recordings of VMN glucose-excited and glucose-inhibited neurons in brain slices from wild-type (WT) mice (A) or nNOS knockout mice (B). The dotted lines represent the resting membrane potential. Glucose concentration changes are schematically displayed below each recording. Downward deflections in whole-cell current-clamp recordings represent the membrane voltage responses to constant hyperpolarizing currents. (C) Table summarizing the number (and %) of VMN glucose-excited (GE), glucose-inhibited (GI), or nonglucose-sensitive (NG) neurons in wild-type or nNOS knockout mice.