Hepatic acetyl CoA links adipose tissue inflammation to hepatic insulin resistance and type 2 diabetes

Abstract

Impaired insulin-mediated suppression of hepatic glucose production (HGP) plays a major role in the pathogenesis of type 2 diabetes (T2D), yet the molecular mechanism by which this occurs remains unknown. Using a novel in vivo metabolomics approach, we show that the major mechanism by which insulin suppresses HGP is through reductions in hepatic acetyl CoA by suppression of lipolysis in white adipose tissue (WAT) leading to reductions in pyruvate carboxylase flux. This mechanism was confirmed in mice and rats with genetic ablation of insulin signaling and mice lacking adipose triglyceride lipase. Insulin's ability to suppress hepatic acetyl CoA, PC activity, and lipolysis was lost in high-fat-fed rats, a phenomenon reversible by IL-6 neutralization and inducible by IL-6 infusion. Taken together, these data identify WAT-derived hepatic acetyl CoA as the main regulator of HGP by insulin and link it to inflammation-induced hepatic insulin resistance associated with obesity and T2D.

Figure 1. Rapid Suppression of Hepatic Glucose Production Rates Is Temporally Associated with a Parallel Rapid Suppression of Lipolysis

(A) Hepatic glucose production measured by the Steele equation (Steele, 1959). Comparisons by ANOVA with Bonferroni’s multiple comparisons test. ****p < 0.0001 versus time 0. (B and C) Plasma non-esterified fatty acid (NEFA) and glycerol concentrations during the hyperinsulinemic-euglycemic clamp. In (A)–(C), n = 6 per time point. (D) Hepatic acetyl CoA concentrations. n = 4 per time point. *p < 0.01, **p < 0.01 versus time 0 by the two-tailed unpaired Student’s t test. (E and F) Whole-body palmitate and glycerol turnover rates. (G–I) Liver acetyl CoA concentrations, hepatic glucose production, and pyruvate carboxylase activity in rats undergoing a hyperinsulinemic-euglycemic clamp with co-infusion of acetate. In (G)–(L), n = 4 per group. (J–L) Plasma glycerol concentrations, whole-body glycerol turnover, and hepatic glucose production in rats undergoing a hyperinsulinemic-euglycemic clamp with co-infusion of glycerol. In (G)–(L), n = 4 per group. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 versus 0 μmol/(kg-min). In all panels, data are mean ± SEM. See also Figure S1.

Figure 2. Normalizing Hepatic Acetyl CoA Concentrations and Whole-Body Glycerol Turnover to Baseline Levels in the Hyperinsulinemic-Euglycemic Clamp Abrogates Insulin’s Ability to Suppress Hepatic Glucose Production

(A) Liver acetyl CoA content. (B) Whole-body glycerol turnover. (C) Plasma glycerol concentrations. (D) Glucose infusion rate. (E) Hepatic glucose production from oxaloacetate (lower bars) and glycerol (upper bars). (F) PC activity. In all panels, ****p < 0.0001 versus basal, §§§§p < 0.0001 versus hyperinsulinemic-euglycemic clamp. n = 6 per group. Data are mean ± SEM. Comparisons by ANOVA with Bonferroni’s multiple comparisons test. See also Figure S2.

Figure 3. Suppression of Lipolysis Results in Decreased Hepatic Acetyl CoA Concentrations and Suppression of Hepatic Glucose Production Independent of Hepatic Insulin Signaling

(A) Protein expression in Akt1 and 2 and Foxo1 knockout mice. (B) Hepatic glucose production. In all panels, black bars, basal; white bars, clamp; and gray bars, clamp+acetate+glycerol. (C) Hepatic acetyl CoA concentrations. (D) Hepatic pyruvate carboxylase activity. In (B)–(D), *p < 0.05, **p < 0.01, ***p < 0.001 versus basal, §p < 0.05, §§p < 0.01, §§§p < 0.001 versus hyperinsulinemic-euglycemic clamp by ANOVA. Data are mean ± SEM of n = 5 per group (basal), 8–9 per group (hyperinsulinemic-euglycemic clamp), and 3–4 per group (hyperinsulinemic-euglycemic clamp + acetate + glycerol). (E) Protein expression in insulin receptor knockdown rats. (F) Hepatic glucose production. Atglistatin-treated rats were given 600 μmol/kg. (G) Liver acetyl CoA. (H) Hepatic PC activity. In (F)–(H), ****p < 0.0001 versus basal, §§§§p < 0.0001 versus hyperinsulinemic-euglycemic clamp by ANOVA. Data are mean ± SEM of n = 6 per group. See also Figure S3.

Figure 4. Four Weeks of High-Fat Feeding Increases Hepatic Glucose Production as a Result of Increased Lipolysis

(A) Hepatic glucose production from oxaloacetate and glycerol. (B and C) Liver acetyl CoA and PC activity. (D and E) Whole-body palmitate and glycerol turnover. In (A)–(E), data are mean ± SEM of n = 6 per group. *p < 0.05, **p < 0.01, ****p < 0.0001 versus chow-fed, basal rats by the two-tailed unpaired Student’s t test. (F) Hepatic glucose production in basal and clamped high-fat-fed adipose-specific ATGL knockout mice. (G and H) Liver acetyl CoA and PC activity. (I and J) Whole-body palmitate and glycerol turnover. In (F)–(J), data are mean ± SEM of n = 8 per group. ***p < 0.001, ****p < 0.0001 versus basal WT, ####p < 0.0001 versus basal ATGL knockouts by the two-tailed unpaired Student’s t test. See also Figures S4 and S5.

Figure 5. Increases in Plasma Interleukin-6 Concentrations Cause Hyperglycemia in High-Fat-Fed Rats

(A and B) IL-6 concentrations in plasma and white adipose tissue macrophages. (C–F) Hepatic glucose production, PC activity, liver acetyl CoA, and whole-body palmitate turn-over in high-fat-fed rats treated with an IL-6-neutralizing antibody. (G and H) Hepatic glucose production, liver PC activity, hepatic acetyl CoA, and whole-body palmitate turnover in control rats treated with or without IL-6 or IL-6 and atglistatin (200 αmol/kg). In all panels, data are mean ± SEM of n = 6 per group. Comparisons by the two-tailed unpaired Student’s t test (A–E) or by ANOVA with Bonferroni’s multiple comparisons test (F–H). *p < 0.05, **p < 0.01, ***p < 0.001 versus controls; §§p < 0.01, §§§p < 0.001 versus chow-fed, IL-6-infused rats. See also Figure S6.

Figure 6. Mice Lacking JNK in Macrophages Are Protected from Diet-Induced Hepatic Insulin Resistance

(A and B) Plasma and WAT macrophage IL-6 concentrations. (C) Glucose infusion rate to maintain euglycemia during a hyperinsulinemic-euglycemic clamp. (D) Liver acetyl CoA concentrations. (E) Hepatic glucose production. (F) Liver PC activity. (G and H) Whole-body palmitate and glycerol turnover. **p < 0.01, ***p < 0.001, ****p < 0.0001 versus basal mice; §p < 0.05, §§§p < 0.001, §§§§p < 0.0001 versus wild-type, clamped mice. Data are mean ± SEM of n = 10 (WT), 8 (m§JNK KO), or 10 (m§JNK KO+IL-6) per group. The groups were compared by ANOVA with Bonferroni’s multiple comparisons test. See also Figure S7.

Figure 7. Insulin-Resistant Obese Adolescents Have Increased Lipolysis and Impaired Suppression of HGP Associated with Increased WAT IL-6 Concentrations

(A and B) Fasting and 2-hr post-challenge plasma glucose. (C) Fasting plasma insulin. (D) HOMA. (E) Hepatic glucose production. Closed bars, basal; open bars, hyperinsulinemic-euglycemic clamp. (F) Fasting NEFA. (G) Suppression of glycerol turnover in the clamp. (H) Mean diameter of large adipocytes. (I) WAT macrophage counts. (J and K) Plasma and WAT IL-6 concentrations. (L) WAT CGI-58 protein. Data are the mean ± SEM of 9–21 controls and 15–39 insulin-resistant subjects. Comparisons were made by the two-tailed unpaired Student’s t test. See also Table S1.