An afferent vagal nerve pathway links hepatic PPARalpha activation to glucocorticoid-induced insulin resistance and hypertension

Abstract

Glucocorticoid excess causes insulin resistance and hypertension. Hepatic expression of PPARalpha (Ppara) is required for glucocorticoid-induced insulin resistance. Here we demonstrate that afferent fibers of the vagus nerve interface with hepatic Ppara expression to disrupt blood pressure and glucose homeostasis in response to glucocorticoids. Selective hepatic vagotomy decreased hyperglycemia, hyperinsulinemia, hepatic insulin resistance, Ppara expression, and phosphoenolpyruvate carboxykinase (PEPCK) enzyme activity in dexamethasone-treated Ppara(+/+) mice. Selective vagotomy also decreased blood pressure, adrenergic tone, renin activity, and urinary sodium retention in these mice. Hepatic reconstitution of Ppara in nondiabetic, normotensive dexamethasone-treated PPARalpha null mice increased glucose, insulin, hepatic PEPCK enzyme activity, blood pressure, and renin activity in sham-operated animals but not hepatic-vagotomized animals. Disruption of vagal afferent fibers by chemical or surgical means prevented glucocorticoid-induced metabolic derangements. We conclude that a dynamic interaction between hepatic Ppara expression and a vagal afferent pathway is essential for glucocorticoid induction of diabetes and hypertension.

Figure 1

Selective hepatic vagotomy reverses glucocorticoid-induced insulin resistance. (A) Operative field for selective nerve interruption. The liver is visible in the left upper quadrant. The esophagus, ventral trunk of the vagus nerve, and the hepatic branch of the vagus are denoted by arrows. (B) Body weight before and after selective hepatic vagotomy or sham surgery. Ppara^+/+ mice were treated with Dex for 5 months (1 mg/kg IP every other day) then underwent either selective hepatic vagotomy or sham surgery. Weight data in grams are presented for 10 mice in each group around the time of the surgery. Weeks on the horizontal axis are not to scale and represent time before and after surgery, which is indicated as “Sx”. (C) Glucose tolerance testing (IP administration of 1 g/kg D-glucose to fasted mice at time 0 followed by blood glucose measurements) in Ppara^+/+ (n=13) and Para^−/− (n=13) mice after 5 months of treatment with Dex. (D) Glucose tolerance testing in Ppara^+/+ mice treated with Dex as in Figure 1C, then randomized to either selective hepatic vagotomy (n=7, open symbols) or sham surgery (n=6, closed symbols). (E) Glucose tolerance testing in Ppara^−/− mice treated with Dex as in Figure 1C, then randomized to either selective hepatic vagotomy (n=5, open symbols) or sham surgery (n=5, closed symbols). (F) Insulin tolerance testing (injection of fasted mice with 0.75 units/kg insulin at time 0 followed by blood glucose measurements) for the animals of Figure 1C. (G) Insulin tolerance testing for the animals of Figure 1G. (H) Insulin tolerance testing for the animals of Figure 1H. Results are presented as mean ± SEM. * P < 0.05 at the same time point by two-tailed, unpaired t test. Panels show results from representative experiments. Similar results were seen in at least 3 independent experiments with different cohorts of mice.

Figure 2. Effects of vagotomy on insulin sensitivity, hepatic gene expression, and reconstitution of PPARα in Pppar^−/− mice

(A-C) Hyperinsulinemic-euglycemic clamp data from Ppara^+/+ mice after chronic treatment with Dex followed by selective hepatic vagotomy (n=5) or sham surgery (n=5) (left side of each panel) and after chronic treatment with saline followed by hepatic vagotomy (n=4) or sham surgery (n=4) (right side of each panel). (D-F) Expression of Ppara, and the known Ppara target genes Acox and Acadam in liver. RNA was analyzed by semi-quantitative RT-PCR. Samples were isolated from Ppara^+/+ mice treated chronically with Dex then subjected to selective hepatic vagotomy (n=4) or sham surgery (n=4). (G) Glucose tolerance testing in Dex-treated mice. Ppara^−/− mice were treated chronically with Dex followed by selective hepatic vagotomy (n=7) or sham surgery (n=7). After recovery from surgery, Ppara^+/+ expression was reconstituted in liver (as documented by RT-PCR, see below) and glucose tolerance testing was performed. (H) Glucose tolerance testing in saline-treated mice. Ppara^−/− mice were treated chronically with saline followed by selective hepatic vagotomy (n=5) or sham surgery (n=4). After recovery from surgery, Ppara^+/+ expression was reconstituted in liver and glucose tolerance testing was performed. Results are presented as mean ± SEM. * indicates P < 0.05 by two-tailed, unpaired t test except for Figure 2C, where ANOVA was used. (I) Tissue survey of gene expression by RT-PCR following treatment with the PPARα adenovirus. Lane 1, water control; lane 2, whole brain; lane 3, brainstem; lane 4, hypothalamus; lane 5, liver; lane 6, heart; lane 7, lung.

Figure 3

Effects of vagotomy on blood pressure. Ppara^+/+ and Ppara^−/−mice were treated with Dex for 5 months (1 mg/kg IP every other day) then systolic (SBP) and diastolic (DBP) blood pressures were determined by tail cuff (left panel, n=13 for each genotype). Mice then underwent either selective hepatic vagotomy or sham surgery, and blood pressure measurements were repeated following recovery in Ppara^+/+ (middle panel, n=14 for vagotomy, n=10 for sham surgery) and Ppara^−/− (right panel, n=14 for vagotomy, n=8 for sham surgery) mice. The blood pressure effects in Ppara^+/+ mice were confirmed invasively (see Supplemental Figure 3). (B) Ppara^+/+ and Ppara^−/−mice were treated with normal saline injections for 5 months then systolic (SBP) and diastolic (DBP) blood pressures were determined by tail cuff (left panel, n=10 for each genotype). These control mice then underwent either selective hepatic vagotomy or sham surgery, and blood pressure measurements were repeated following recovery in Ppara^+/+ (middle panel, n=5 for vagotomy, n=5 for sham surgery) and Ppara^−/− (right panel, n=5 for vagotomy, n=5 for sham surgery) mice. (C) Ppara^−/− mice were treated chronically with Dex followed by selective hepatic vagotomy (n=7) or sham surgery (n=7). After recovery from surgery, Ppara^+/+ expression was reconstituted in liver (as documented by RT-PCR), blood pressures were measured by tail cuff (shown) and confirmed invasively (see Supplemental Figure 3). (D) Ppara^−/− mice were treated chronically with saline followed by selective hepatic vagotomy (n=5) or sham surgery (n=5). After recovery from surgery, Ppara^+/+ expression was reconstituted in liver and blood pressures were measured. Results are presented as mean ± SEM. * P < 0.05 by two-tailed, unpaired t test.

Figure 4

Selective peripheral afferent vagal denervation reverses Dex-induced insulin resistance. Dex-treated Ppara^+/+ mice underwent either capsaicin or vehicle application to the hepatic vagal branch. Electron microscopy (EM) of hepatic vagal nerve sections (Panels A-D) and physiological studies (Panels E-G) were performed 10 days later. In panels A-D, arrows represent unmyelinated fibers and arrowheads myelinated fibers. (A) EM assessment of the vagus in vehicle-treated mice. (B) EM evaluation from the posterior vagal trunk of an animal subjected to capsaicin treatment of the hepatic vagal branch. (C, D) EM of hepatic vagal nerves from two different capsaicin-treated animals. (E) Food intake in grams per day from four animals in each group. (D) Body weight after capsaicin or vehicle treatment. Weight data in grams are presented for 5 mice in each group. The horizontal axis represents days after the procedure. (E) Glucose tolerance testing in capsaicin-treated (n=5) and vehicle-treated (n=5) mice. Results are presented as mean ± SEM. * P < 0.05 at the same time point by two-tailed, unpaired t test. Panels show results from representative experiments. Similar results were seen in 2 independent experiments with different cohorts of mice.

Figure 5

Central afferent vagal nerve sectioning reverses Dex-induced hypertension and insulin resistance. Dex-treated Ppara^+/+ mice were randomly assigned to either selective section of the vagal afferent fibers in the brainstem or a sham operation. Panels A-C represent views of the operative field with the area enclosed by the rectangle in A expanded in B, and the rectangle in B expanded in C. (D) Body weight in grams is presented for 3 mice in each group. The horizontal axis represents weeks after the surgical procedure. (E) Tail-cuff systolic (SBP) and diastolic (DBP) blood pressure were obtained three weeks after the procedure (n=3 for each group). (F) Glucose tolerance testing three weeks after central afferent vagal sectioning (n=3) and the sham operation (n=3). Results are presented as mean ± SEM. * P < 0.05 at the same time point by two-tailed, unpaired t test.