Control of physiologic glucose homeostasis via hypothalamic modulation of gluconeogenic substrate availability

Abstract

Objectives: The brain mobilizes glucose in emergency situations such as hypoglycemia as well as during day-to-day physiology such as fasting. While most hypothalamic neuronal populations that contribute to glucose mobilization also contribute to other aspects of metabolism, neurons in the ventromedial nucleus of the hypothalamus that express the cholecystokinin b receptor (VMH^Cckbr neurons) support glucose production during hypoglycemia without controlling energy homeostasis. However, their role in day-to-day glucose physiology and the mechanisms they engage to support glucose mobilization is unclear.

Methods: We used continuous glucose monitoring in mice with chronically silenced VMH^Cckbr neurons to establish whether these neurons are required during day-to-day glucose homeostasis. Tetanus-toxin based chronic silencing and acute optogenetic activation were followed by analysis of hepatic glucose metabolism and white adipose tissue lipolysis.

Results: We found that VMH^Cckbr neurons support glucose homeostasis during short fasts and contribute to gluconeogenic substrate mobilization and lipolysis. VMH^Cckbr neurons mobilize glucose without depleting hepatic glycogen or increasing gluconeogenic gene expression, but instead mobilize glycerol in a β3-adrenergic receptor (β3-AR)-dependent manner. Restoring glycerol availability following VMH^Cckbr neuron silencing restores glucose. Finally, acute activation of VMH^Cckbr neurons mobilizes additional gluconeogenic substrates beyond glycerol.

Conclusions: VMH^Cckbr neurons represent a distinct subset of glucose-mobilizing VMH neurons that support physiologic glucose homeostasis, likely through control of β3-AR-mediated gluconeogenic substrate mobilization and lipolysis. The presence of different glucose-mobilizing neuronal populations that engage distinct mechanisms in a context-dependent manner may provide the brain with flexibility to coordinate the appropriate glycemic response to different circumstances.

Keywords: Glucose metabolism; Hepatic glucose production; Hypothalamus; Lipolysis; Sympathetic nervous system.

Figure 1

Silencing VMH^Cckbr neurons decreases glucose during short fasts. We injected Cckbr^Cre mice with AAV-DIO-TetTox-EGFP bilaterally into the VMH (n = 11) or AAV-DIO-GFP as a control (n = 10). (A) Blood glucose over a 24 h period was monitored via continuous glucose monitoring in a subset of mice (n = 4 Cckbr^TetTox and 5 control) (ZT = Zeitgeber Time, TT = Tetanus Toxin). Following a 4 h fast we measured (B) plasma hormones, (C) hepatic gluconeogenic gene expression, and (D) liver glycogen content. Data are plotted as mean +/− SEM, analyzed by unpaired Student's t-test. Differences in blood glucose by CGM was analyzed in 4-hr circadian windows by 2-way ANOVA. ∗p < 0.05.

Figure 2

Activating VMH^Cckbr neurons mobilizes glucose through mechanisms distinct from activation of the entire VMH. Cckbr^Cre mice were injected with AAV-DIO-ChR2-eYFP unliterally into the VMH followed by optogenetic fiber placement to permit optogenetic activation of VMH^Cckbr neurons. Mice were fasted for 2 h during the early light cycle prior to testing. Shown is (A) the glycemic response to light or no light control stimulation (n = 19), (B) mRNA expression of hepatic gluconeogenic genes was quantified following 30 min control or light delivery (n = 6 light off, 8 light on) and (C) liver glycogen content from tissue collected after 30 min of light or control stimulation (n = 5 no light, 8 light). (D) A separate cohort of mice were treated with CP91,149, a glycogenolysis inhibitor, or vehicle at the start of the fasting period and glucose was measured before and 30 min after optogenetic stimulation (n = 5); the right-hand panel shows the percent change for light-stimulated vs no light conditions for each treatment. Data are plotted as mean SEM. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < 0.0001, by paired Student's t-test (A, D) and unpaired Student's t-test (B, C, D (t = 0 comparison)).

Figure 3

Silencing VMH^Cckbr neurons impairs glycerol mobilization and lipolysis. (A) Gluconeogenic substrates were measured following a 4 h fast in Cckbr^TetTox (n = 7) and GFP control mice (n = 5). (B) Plasma NEFA concentration was measured following a 4 h fast (n = 7–8). (C) Animals underwent a glycerol appearance assay with infusion of [³H]-labelled glycerol and glycerol appearance rate was calculated (n = 5–7). (D) Plasma glycerol from Cckbr^Cre mice injected with AAV-DIO-ChR2-eYFP unliterally into the VMH followed by optogenetic fiber placement in response to light (activation) or no light (control) (n = 8). Data are plotted as mean +/− SEM. ∗p < 0.05, ∗∗∗p < 0.001, by unpaired Students t-test.

Figure 4

β3-adrenergic receptor blockade prevents VMH^Cckbr-mediated glycerol and ketone mobilization. (A) Plasma glycerol concentration in Cckbr^Cre mice injected with AAV-DIO-ChR2-eYFP unliterally into the VMH followed by optogenetic fiber placement treated with vehicle (V) or the β3-AR antagonist SR-59230A (SR) at baseline and following 60 min of optogenetic activation (light) or control (no light). (B) Blood ketone concentration measured at 30 min intervals during light stimulation in mice pre-treated with β3-AR antagonist SR-59230A or vehicle (n = 9). Data are plotted as individuals (A) or mean +/− SEM (B) and analyzed by paired Student's t-test. ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

Figure 5

Restoring glycerol availability in Cckbr^TetTox mice restores blood glucose. Ccbkr^TetTox and control mice were injected with (A) pyruvate (2 g/kg) or (B) glycerol (2 g/kg) following a 4-hour fast and blood glucose was measured at the indicated times. (C) Glycerol, (D) glycerol appearance rate, and (E) blood glucose were measured in Cckbr^TetTox and control mice following a 4-hour fast at 0 and 10 min after β3-agonist CL316,243 (CL) i.v. infusion (n = 5). Data are plotted as mean +/− SEM. ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001 by unpaired Student's t-test (A–D) or paired Student's t-test (E).

Figure 6

Acute VMH^Cckbr neuron activation mobilizes multiple gluconeogenic substrates, (A) Blood glucose measured at 30 min intervals during optogenetic stimulation of VMH^Cckbr neurons in mice pre-treated with β3-AR antagonist SR-59230A or vehicle (n = 9). (B) Plasma concentration of plasma lactate, pyruvate, glycerol and amino acids following 30 min optogenetic stimulation of VMH^Cckbr neurons (ON) compared to no light control (OFF) (n = 10–15). Data are plotted as individual values (left panel) and average change from baseline (mean +/− SEM, right panel). ∗p < 0.05, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001 by paired Student's t-test (A, B (left)) or unpaired Student's t-test (B, right).