Local and systemic insulin resistance resulting from hepatic activation of IKK-beta and NF-kappaB

Abstract

We show that NF-kappaB and transcriptional targets are activated in liver by obesity and high-fat diet (HFD). We have matched this state of chronic, subacute 'inflammation' by low-level activation of NF-kappaB in the liver of transgenic mice, designated LIKK, by selectively expressing constitutively active IKK-b in hepatocytes. These mice exhibit a type 2 diabetes phenotype, characterized by hyperglycemia, profound hepatic insulin resistance, and moderate systemic insulin resistance, including effects in muscle. The hepatic production of proinflammatory cytokines, including IL-6, IL-1beta and TNF-alpha, was increased in LIKK mice to a similar extent as induced by HFD in in wild-type mice. Parallel increases were observed in cytokine signaling in liver and mucscle of LIKK mice. Insulin resistance was improved by systemic neutralization of IL-6 or salicylate inhibition of IKK-beta. Hepatic expression of the IkappaBalpha superrepressor (LISR) reversed the phenotype of both LIKK mice and wild-type mice fed an HFD. These findings indicate that lipid accumulation in the liver leads to subacute hepatic 'inflammation' through NF-kappaB activation and downstream cytokine production. This causes insulin resistance both locally in liver and systemically.

Figures

Figure 1

IKK-β and NF-κB activities in liver. (a) NF-κB and (b) IKK-β activities are expressed as fold differences (n = 6, *P < 0.05; **P < 0.01). (c) Schematic map of the transgene construct showing Flag-tagged IKK-β S177E,S181E subcloned into exon 2 of the Alb1 promoter-driven truncated β-globin gene vector. (d) Transgene expression and (e) total amounts of IKK-β in skeletal muscle (Mus), kidney (Kid), white adipose tissue (Wat), liver (Liv) and pancreas (Pan). (f) IKK-β and (g) NF-κB activities in liver (n = 4–6, **P < 0.01). (h) Hematoxylin and eosin–stained sections of liver. Arrows indicate central veins. Scale bars, 50 μm.

Figure 2

Carbohydrate metabolism. (a) Fasting insulin levels (n = 8–12; AUC, area under curve). (b) Hyperinsulinemic-euglycemic clamps were performed in 12–15-week-old male mice (n = 10, *P < 0.05). (c,d) Glucose tolerance tests were performed in 12-week-old male mice (n = 6). (c) Glucose concentrations or (d) insulin resistance index (IRI = glucose concentration (mmol/L) × insulin concentration (mU/L) ÷ 22.5) were plotted versus time (left) and as AUC (right, per 120 min). *P < 0.05, **P < 0.01. (e) Hepatic glucose production and glycogen synthesis determined during hyperinsulinemic-euglycemic clamps (n = 10; **P < 0.01). (f) GSK3β phosphorylation in liver after hyperinsulinemic-euglycemic clamps. (g) mRNA expression levels for gluconeogenic enzymes determined using real-time RT-PCR (*P < 0.05, **P < 0.01). (h) Glycogen synthesis and glucose uptake in skeletal muscle during hyperinsulinemic-euglycemic clamps (n = 10, *P < 0.05).

Figure 3

Insulin signaling in liver and skeletal muscle of LIKK mice. Fasting mice were injected with insulin (5.0 mU/g) or saline. Proteins were immunoprecipitated (IP) from (a,b) liver or (c,d) skeletal muscle (gastrocnemius) immunoblotted (IB) with the indicated antibodies. (b,d) Specific phosphorylation: open bars, without insulin; closed bars, with insulin; n = 3; *P < 0.05, **P < 0.01. Ins, insulin; IR, insulin receptor.

Figure 4

Cytokine signaling in LIKK mice. (a) Western blots of phosphorylated Stat-3 or Stat-3 in liver. (b) Circulating levels of TNF-α, IL-1β and IL-6 (n = 4–8; *P < 0.05, ** P < 0.01). (c,d) mRNA levels in skeletal muscle determined by real-time RT-PCR. NF-κB activity measured by specific DNA binding (n = 6; * P < 0.05). (e) Glucose tolerance tests 2–3 weeks after treatment with neutralizing antibody to IL-6 (αIL6) or control IgG (n = 6; insulin resistance index, IRI = glucose concentration (mmol/L) × insulin concentration (mU/L) ÷ 22.5). (f) Phosphorylated Stat-3 and Stat-3 in livers from mice treated with antibodies to IL-6 (αIL6) or control IgG. mRNA levels in liver (g) and gastrocnemius muscle (h) determined by real-time RT-PCR (*P < 0.05, ** P < 0.01). Il1b encodes IL-1β; Il1R1 encodes IL-1R1.

Figure 5

Reversal of insulin resistance and inflammation. (a) Mice (16-week-old males; n = 6–8) treated orally with high-dose salicylate were subjected to glucose tolerance testing (insulin resistance index = glucose concentration (mmol/L) × insulin concentration (mU/L) ÷ 22.5). (b,c) Binding of NF-κB to DNA in liver and skeletal muscle of control and sodium salicylate-treated (NaSal) mice. (d) Doubly transgenic LIKK × LISR mice were subjected to glucose tolerance testing (12-week-old males; n = 5 per group). (e) Mice fed a HFD for 3 months were subjected to glucose tolerance testing (n = 6–7 per group). mRNA levels of monocyte-macrophage proteins in liver of (f) chow-fed wild-type and LIKK mice and (g) wild-type and LISR mice fed chow or HFD (P < 0.05, **P < 0.01).