Exogenous glucose administration impairs glucose tolerance and pancreatic insulin secretion during acute sepsis in non-diabetic mice

Abstract

Objectives: The development of hyperglycemia and the use of early parenteral feeding are associated with poor outcomes in critically ill patients. We therefore examined the impact of exogenous glucose administration on the integrated metabolic function of endotoxemic mice using our recently developed frequently sampled intravenous glucose tolerance test (FSIVGTT). We next extended our findings using a cecal ligation and puncture (CLP) sepsis model administered early parenteral glucose support.

Methods: Male C57BL/6J mice, 8-12 weeks, were instrumented with chronic indwelling arterial and venous catheters. Endotoxemia was initiated with intra-arterial lipopolysaccharide (LPS; 1 mg/kg) in the presence of saline or glucose infusion (100 µL/hr), and an FSIVGTT was performed after five hours. In a second experiment, catheterized mice underwent CLP and the impact of early parenteral glucose administration on glucose homeostasis and mortality was assessed over 24 hrs.

Measurements: AND MAIN RESULTS: Administration of LPS alone did not impair metabolic function, whereas glucose administration alone induced an insulin sensitive state. In contrast, LPS and glucose combined caused marked glucose intolerance and insulin resistance and significantly impaired pancreatic insulin secretion. Similarly, CLP mice receiving parenteral glucose developed fulminant hyperglycemia within 18 hrs (all > 600 mg/dl) associated with increased systemic cytokine release and 40% mortality, whereas CLP alone (85 ± 2 mg/dL) or sham mice receiving parenteral glucose (113 ± 3 mg/dL) all survived and were not hyperglycemic. Despite profound hyperglycemia, plasma insulin in the CLP glucose-infused mice (3.7 ± 1.2 ng/ml) was not higher than sham glucose infused mice (2.1 ± 0.3 ng/ml).

Conclusions: The combination of parenteral glucose support and the systemic inflammatory response in the acute phase of sepsis induces profound insulin resistance and impairs compensatory pancreatic insulin secretion, leading to the development of fulminant hyperglycemia.

Figure 1. Experimental timelines.

(A) Chronically catheterized mice (femoral artery and vein) were given three days to recover from surgery and infused intravenously (iv) with either saline or glucose for 24 hours prior to intra-arterial administration of either lipopolysaccharide (LPS; 1 mg/kg) or vehicle. Five hours after LPS or vehicle administration a frequently sampled intravenous glucose tolerance test (FSIVGTT) was performed over a two hour period. An iv glucose bolus of 1 g/kg D50 was given over approximately 15 sec (▲). Subsequently, multiple samples of either blood glucose and plasma insulin (Δ) or glucose alone (↑) were taken from the arterial catheter at the times identified. Red blood cells were spun down and re-infused into the mouse throughout the protocol to avoid anemia. (B) Mice underwent cecal ligation and puncture (CLP) immediately after chronic catheterization of the femoral artery and vein. Animals were infused intravenously with either saline or glucose upon completion of surgery and followed for 24 hrs. Blood glucose and plasma insulin (Δ) or blood glucose alone (↑) were sampled from the arterial catheter at the times identified.

Figure 2. Physiological variables.

Mean ± SEM for (A) body weight, (B) food intake over 24 hrs prior to lipopolysaccharide (LPS; 1 mg/kg) administration (note: glucose infused mice consumed fewer calories as the glucose accounted for approximately 40% of their daily calorie intake), (C) pre-FSIVGTT blood glucose, and (D) plasma insulin 5 hrs after either LPS or vehicle (Veh) administration in mice infused with either saline (Sal) or glucose (Glu) at 100 µL/hr for 29 hrs. Statistical differences were determined by two-way ANOVA with differences between individual means determined by post-hoc Bonferroni tests.

Figure 3. Hemodynamic profiles before and after intra-arterial endotoxin.

Change in mean arterial blood pressure over time in four groups of animals receiving either (1) intra-venous saline infusion and vehicle (Veh Sal), (2) intra-venous saline infusion intra-arterial lipopolysaccharide (1 mg/kg, LPS Sal), (3) intra-venous glucose infusion and vehicle (Veh Glu), (4) intra-venous glucose infusion and LPS (1 mg/kg, LPS Glu). Time scale is adjusted to indicate effects of intra-venous saline and glucose infusion, LPS administration, and the frequently sampled intravenous glucose tolerance test (FSIVGTT).

Figure 4. Frequently sampled intravenous glucose tolerance test data.

Glucose disposal curves in response to an iv injection of 1 g/kg D50 administered over 15 sec at time t = 0 are depicted for (A) mice receiving iv saline infusion and vehicle (Veh Sal, open diamonds) or iv saline infusion and intra-arterial lipopolysaccharide (1 mg/kg, LPS Sal, closed diamonds) and (B) iv glucose infusion and vehicle (Veh Glu, open triangles) or mice receiving iv glucose infusion and LPS (1 mg/kg, LPS Glu, closed triangles). The corresponding insulin response curves are shown for (C) Veh Sal and LPS Sal and (D) Veh Glu and LPS Glu groups.

Figure 5. Minimal model data from frequently sampled intravenous glucose tolerance test.

Data are shown for (A) glucose tolerance (area under the glucose curve; AUCg), (B) insulin sensitivity (Si), (C) acute insulin response to glucose (AIRg), and (D) Disposition Index (DI = Si *AIRg) for four groups of animals receiving either iv saline infusion and vehicle (Veh Sal, open diamonds), iv saline infusion and intra-arterial lipopolysaccharide (1 mg/kg, LPS Sal, closed diamonds), iv glucose infusion and vehicle (Veh Glu, open triangles), or iv glucose infusion and LPS (1 mg/kg, LPS Glu, closed triangles). Group medians are indicated by gray lines. Statistical differences were determined by Kruskal-Wallis test with differences between individual means determined by post-hoc Wilcoxon rank-sum tests corrected for multiple comparisons using the Bonferroni method.

Figure 6. Metabolic and hemodynamic profiles after cecal ligation and puncture.

Scatter plot of individual data sets for (A) 18 hr blood glucose, (B) 18 hr plasma insulin and (C) mean arterial blood pressure for mice that have undergone either CLP or sham surgery and infused with iv saline (Sal) or glucose (Glu) at 100 µL/hr for 24 hrs post-operatively (Sham Sal, open diamonds, Sham Glu, open triangles, CLP Sal, closed diamonds, or CLP Glu, closed triangles). Mean arterial blood pressure was averaged over consecutive periods of 8 hrs, 12 hrs, and 4 hrs. Panel D depicts bacterial colony forming units (CFU) in blood extracted from mice 24 hrs after CLP and subject to post-operative saline or glucose infusion. Statistical differences were determined by Kruskal-Wallis test with differences between individual means determined by post-hoc Wilcoxon rank-sum tests corrected for multiple comparisons using the Bonferroni method.

Figure 7. Inflammatory cytokine profiles after cecal ligation and puncture.

Mean ± SEM for Interleukin-1β (IL-1β; A), Interleukin-6 (IL-6; B), Tumor necrosis factor-α (TNF-α; C), and Interleukin-10 (IL-10; D) measured in plasma 18 hrs after either CLP or sham surgery and infusion with iv saline (Sal) or glucose (Glu) at 100 µL/hr post-operatively. Statistical differences were determined by two-way ANOVA after log transformation where appropriate with differences between individual means determined by post-hoc Bonferroni tests.