Still Excited, but Less Aroused-The Effects of Nutritional Ketosis on Epinephrine Response and Hypothalamic Orexin Neuron Activation Following Recurrent Hypoglycemia in Diabetic Rats

Abstract

Hypoglycemia-associated autonomic failure (HAAF) is a serious, life-threatening complication of intensive insulin therapy, particularly in people with type 1 diabetes. The ketogenic diet is reported to beneficially affect glycemic control in people with type 1 diabetes, however its effects on the neurohormonal counterregulatory response to recurrent hypoglycemia and HAAF development are understudied. In this study we used Sprague Dawley rats to establish a HAAF model under non-diabetic and streptozotocin (STZ)-induced diabetic conditions and determined how nutritional ketosis affected the neurohormonal counterregulation and the activity of energy-sensing orexin (OX) neurons. We found that antecedent hypoglycemia diminished the sympathoexcitatory epinephrine response to subsequent hypoglycemia in chow-fed non-diabetic rats, but this did not occur in STZ-diabetic animals. In all cases a ketogenic diet preserved the epinephrine response. Contrary to expectations, STZ-diabetic keto-fed rats showed reduced OX activity in the recurrent hypoglycemia group, which did not occur in any other group. It is possible that the reduced activation of OX neurons is an adaptation aimed at energy conservation accompanied by diminished arousal and exploratory behaviour. Our data suggests that while a ketogenic diet has beneficial effects on glycemia, and epinephrine response, the reduced activation of OX neurons could be detrimental and warrants further investigation.

Keywords: STZ-diabetes; counterregulatory response; epinephrine; insulin-induced hypoglycemia; ketogenic diet; ketosis; orexin neurons; rat model.

Figure 1

The effects of three weeks of nutritional ketosis of bodyweight, blood glucose (BG) and β-hydroxybutyrate (BHB) of non-diabetic and STZ-diabetic rats. Diabetes was induced with a single i.p. injection of 60 mg/kg STZ (citrate buffer was used in non-diabetic controls), followed by 3 weeks of dietary intervention. (A) Significantly lower bodyweight in ketogenic diet-fed non-diabetic (Non-D Keto) and diabetic (STZ-D Keto) rats compared to respective chow-fed controls; (B) Non-D Keto rats had lower BG and higher BHB concentrations than chow fed controls; (C) Keto diet did not affect BG in STZ-diabetic rats, but significantly elevated BHB; (D) No differences in baseline BHB of Non-D Keto and STZ-D Keto rats were observed; (E) Baseline BHB inversely correlated with BG in Non-D Keto rats; (F) BHB did not significantly correlate with weight in Non-D Keto rats; (G) Highest BHB concentrations in STZ-D Keto rats corresponded to the highest BG; (H) A strong inverse correlation between bodyweight and BHB was detected in STZ-D Keto rats. Data were analysed by Mann–Whitney tests (A–C), Welch’s t-test (D), Spearman’s correlation (E,F). Data shown as mean ± SEM, * p < 0.05, **** p < 0.0001.

Figure 2

Streptozotocin (STZ) induces hyperglycemia and pancreatic damage in chow and keto-fed rats. (A) A single 60 mg/kg dose of STZ, administered i.p. significantly increased baseline BG levels in both Chow and Keto-fed rats; insulin supplementation decreased BG in the STZ-D Chow rats to the same level as ketogenic diet alone, without insulin supplementation, in STZ-D Keto group; (B) The ratio of insulin (β-cells)/glucagon (α-cells) producing cells in pancreatic islets of all STZ-D groups (No Hypo, 1×-Hypo and 3×-Hypo) of chow-fed and keto-fed rats was significantly reduced compared to vehicle (citrate) controls (Non-D Chow and Non-D Keto, respectively). Data were analysed by one-way ANOVA with Holm-Šídák multiple comparisons test and presented as mean ± SEM, *** p < 0.001, **** p < 0.0001.

Figure 3

Changes in pancreatic islet morphology in STZ-injected rats. 5-µm sections of the pancreas were stained for insulin (green) and glucagon (purple); representative images are shown (scale bar 100 µm). (A,B) Non-diabetic chow-fed control; (C,D) Non-diabetic keto-fed control; (E,F) STZ-treated Chow No Hypo control; (G,H) STZ-treated Keto No Hypo Control; (I,J) STZ-treated Chow single hypoglycemia; (K,L) STZ-treated Keto single hypoglycemia; (M,N) STZ-diabetic Chow recurrent hypoglycemia; (O,P) STZ-diabetic Keto recurrent hypoglycemia.

Figure 4

Ketogenic diet differentially affects hypoglycemia counterregulatory hormones in non-diabetic and diabetic rats. (A) Recurrent hypoglycemia (3×-Hypo) reduces epinephrine counterregulatory response in non-diabetic chow-fed rats; no reduction is observed in non-diabetic keto-fed rats; (B) Epinephrine counterregulatory response to recurrent hypoglycemia (3×-Hypo) is preserved in STZ-diabetic chow- and keto-fed rats; (C) Single (1×-Hypo) or recurrent (3×-Hypo) hypoglycemia did not induce glucagon counterregulatory response in non-diabetic chow-fed animals, but significantly increase its release in non-diabetic keto-fed rats; (D) No changes in glucagon concentration were detected in STZ-diabetic animals. Log-transformed concentrations of epinephrine and glucagon (pg/mL) were analysed by one-way ANOVA with Holm-Šídák multiple comparisons test. Data are mean ± SEM, * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Figure 5

Baseline BG and BHB correlate with post-hypoglycemic epinephrine. Plasma epinephrine and glucagon concentrations were measured in animals subjected to a single (1×-Hypo) or recurrent (3×-Hypo) insulin injections; in the 3×-Hypo groups epinephrine and glucagon were measured after the final insulin injection. (A) In non-diabetic keto-fed rats baseline BG inversely correlated with plasma epinephrine, but not glucagon, measured 2 h after insulin injection (n = 12); (B) Direct correlation was detected between baseline BHB and epinephrine in non-diabetic keto-fed rats, and no correlation with glucagon; (C) Baseline BG did not correlate with epinephrine or glucagon in STZ-diabetic keto-fed rats (n = 10); (D) Baseline BHB significantly correlated with glucagon, but not epinephrine in STZ-diabetic keto-fed rats; (E) Baseline BG was not associated with changes in epinephrine or glucagon in non-diabetic chow-fed rats (n = 10) or (F) STZ-diabetic chow-fed rats (n = 9). Concentrations of epinephrine ang glucagon (pg/mL) were log-transformed for analysis. Data analysed by Spearman’s correlation. Datapoints are individual animals.

Figure 6

Recurrent hypoglycemia attenuates OX neurons activation in STZ-diabetic keto-fed rats. (A) The proportion of Fos-positive OX neurons did not change with hypoglycemic challenge in any of the non-diabetic groups of rats; (B) Single (1×-Hypo) hypoglycemia increased the proportion of activated OX neurons in STZ-diabetic keto-fed rats compared to the No Hypo control and recurrent hypoglycemia (3× Hypo) attenuated this increase; no differences in OX activity were detected in STZ-diabetic chow-fed rats. Data analysed by one way ANOVA with Holm–Šídák multiple comparisons test. Data are mean ± SEM, * p < 0.05.