Glucose Activates Vagal Control of Hyperglycemia and Inflammation in Fasted Mice

Abstract

Sepsis is a leading cause of death in hospitalized patients. Many experimental treatments may have failed in clinical trials for sepsis, in part, because they focused on immune responses of healthy animals that did not mimic the metabolic settings of septic patients. Epidemiological studies show an association between metabolic and immune alterations and over 1/3 of septic patients are diabetic, but the mechanism linking these systems is unknown. Here, we report that metabolic fasting increased systemic inflammation and worsened survival in experimental sepsis. Feeding and administration of glucose in fasted mice activated the vagal tone without affecting blood pressure. Vagal stimulation attenuated hyperglycemia and serum TNF levels in sham but only hyperglycemia in splenectomized mice. Vagal stimulation induced the production of dopamine from the adrenal glands. Experimental diabetes increased hyperglycemia and systemic inflammation in experimental sepsis. Fenoldopam, a specific dopaminergic type-1 agonist, attenuated hyperglycemia and systemic inflammation in diabetic endotoxemic mice. These results indicate that glucose activates vagal control of hyperglycemia and inflammation in fasted septic mice via dopamine.

Figure 1

Fasting increases inflammation and mortality in endotoxemia. Mice were (control) fed ad libitum, (Fasted) fasted for 24 h, or (F/fed) fasted for 20 h and fed for 4 h before challenged with LPS (10 mg/kg; i.p.). Serum levels of TNF (a), IL6 (b), IFNγ (c), HMGB1 (d), TGFβ1 (e) and IL10 (f) were analyzed at the indicated time points. *p < 0.05 vs F/fed (n = 4/group, two-way ANOVA). (g) Blood creatinine levels at 24 h post-LPS. ⁺p < 0.05 vs F/fed (n = 4/group, one-way ANOVA). (h) Kaplan-Meier survival analyses of mice in endotoxemia (LPS 10 mg/kg; i.p.; ^§p < 0.05 vs. F/fed, n = 20/group, Survival Log-rank test).

Figure 2

Glucose activates the vagus nerve. (a) Fasted or fasted/fed (F/fed) mice were challenged with LPS and blood glucose levels were analyzed at different time points. *p < 0.05 vs F/fed (n = 3/group, two-way ANOVA). (b) Fasted animals received saline solution with glucose(0, 0.2 or 2 mg; i.v.) and the vagal nerve activity and arterial blood pressure were recorded (n = 2-3/group for 3 sets of experiments performed for each of the treatments). (c,d) Fasted mice underwent sham or surgical (c) vagotomy (VGX), or (d) splenectomy (SPX) at 30 h or 3 days before LPS, respectively. Animals were treated with saline alone or with glucose (2 mg/Kg; i.p.), and serum TNF levels were analyzed at 1.5 h post-LPS. ⁺p < 0.05 vs LPS (n = 3/group; One-way ANOVA).

Figure 3

Vagal stimulation attenuates hyperglycemia by inducing insulin. (a) Mice were challenged with LPS and received sham or vagal stimulation (VS; 10 Hz, 60 min). Blood glucose levels were analyzed at different time points. *p < 0.05 vs sham (n = 4/group; Two-way ANOVA). (b,c) Mice underwent sham or surgical splenectomy (SPX) 3 days before LPS. Animals underwent control surgery or vagal stimulation (VS), and serum TNF at 1.5 h (b); ⁺p < 0.05 vs LPS (n = 3/group; One-way ANOVA), or blood glucose levels (c) were analyzed. *p < 0.05 vs LPS (n = 4/group; Two-way ANOVA). (d) Fasted mice received sham or vagal stimulation (VS), and serum insulin levels were analyzed at the indicated time points. *p < 0.05 vs LPS (n = 4/group; Two-way ANOVA). (e,f) Fasted mice underwent sham or surgical pancreatectomy (PX) 1 day before LPS, underwent sham or vagal stimulation, and blood insulin (e) or glucose (f) levels were analyzed at 1.5 h post-LPS. ⁺p < 0.05 vs LPS (n = 4/group; One-way ANOVA).

Figure 4

Dopamine controls hyperglycemia in experimental sepsis. (a–c) Fasted animals underwent (a) sham or (a,c) surgical adrenalectomy (ADX) 3 days before LPS. Then, animals underwent control or vagal stimulation (VS; 10 Hz, 15 min) right before the LPS challenge. (a) Blood glucose or (b,c) catecholamines (dopamine(DA), norepinephrine (NE) or epinephrine (E) were analyzed at 1.5 h post-LPS. (d–f) Fasted mice were treated with vehicle (control), dopamine (DA; 10 mg/kg/dose; i.p.) or fenoldopam (Fen; 10 mg/kg/dose; i.p.) at 6 and 1 h before the LPS challenge. (d,e) Blood glucose and (f) serum TNF levels were analyzed at the indicated time points. (g,h) All mice received streptozotocin (STZ; 40 mg/kg) and were treated with vehicle (control) or fenoldopam (Fen; 10 mg/kg/dose; i.p.) at 6 and 1 h before LPS. Blood glucose (g) and TNF (h) were analyzed at the indicated time points. (a,f) ⁺p < 0.05 vs LPS (n = 4/group; One-way ANOVA); (b,c) ^#p < 0.05 vs LPS (n = 3/group; Student’s t-test); (d,e,g,h) *p < 0.05 vs control (n = 4/group, Two-way ANOVA).