Insulin resistance in non-obese subjects is associated with activation of the JNK pathway and impaired insulin signaling in skeletal muscle

Abstract

Background: The pathogenesis of insulin resistance in the absence of obesity is unknown. In obesity, multiple stress kinases have been identified that impair the insulin signaling pathway via serine phosphorylation of key second messenger proteins. These stress kinases are activated through various mechanisms related to lipid oversupply locally in insulin target tissues and in various adipose depots.

Methodology/principal findings: To explore whether specific stress kinases that have been implicated in the insulin resistance of obesity are potentially contributing to insulin resistance in non-obese individuals, twenty healthy, non-obese, normoglycemic subjects identified as insulin sensitive or resistant were studied. Vastus lateralis muscle biopsies obtained during euglycemic, hyperinsulinemic clamp were evaluated for insulin signaling and for activation of stress kinase pathways. Total and regional adipose stores and intramyocellular lipids (IMCL) were assessed by DXA, MRI and (1)H-MRS. In muscle of resistant subjects, phosphorylation of JNK was increased (1.36±0.23 vs. 0.78±0.10 OD units, P<0.05), while there was no evidence for activation of p38 MAPK or IKKβ. IRS-1 serine phosphorylation was increased (1.30±0.09 vs. 0.22±0.03 OD units, P<0.005) while insulin-stimulated tyrosine phosphorylation decreased (10.97±0.95 vs. 0.89±0.50 OD units, P<0.005). IMCL levels were twice as high in insulin resistant subjects (3.26±0.48 vs. 1.58±0.35% H(2)O peak, P<0.05), who also displayed increased total fat and abdominal fat when compared to insulin sensitive controls.

Conclusions: This is the first report demonstrating that insulin resistance in non-obese, normoglycemic subjects is associated with activation of the JNK pathway related to increased IMCL and higher total body and abdominal adipose stores. While JNK activation is consistent with a primary impact of muscle lipid accumulation on metabolic stress, further work is necessary to determine the relative contributions of the various mediators of impaired insulin signaling in this population.

Figure 1. JNK activation in muscle from insulin resistant, non-obese subjects accompanies reduced insulin signaling and enhanced phosphorylation of IRS-1 on Serine³¹².

Solubilized muscle biopsies obtained prior to insulin infusion were subject to SDS-PAGE, followed by Western blotting with phospho-specific antibodies against JNK Thr¹⁸³/Tyr¹⁸⁵ (A), p38 MAPK Thr¹⁸⁰/Tyr¹⁸² (B), and IKKβ Ser¹⁷⁷ (C). The predominant band (55 kD) recognized by the anti-JNK antibody is shown and was used for quantification. Optical density values were normalized to the content values for each protein as determined by Western blots for JNK, p38 MAPK and IKKβ, and are presented as mean ± SEM of the ratios of phospho- to total protein. *P<0.05 RES vs. SEN groups. (D) Muscle biopsies obtained after 2 hours of insulin infusion were solubilized and assayed for phosphorylation state of key second messengers of insulin signal transduction. Phosphorylation of Akt on Thr³⁰⁸ was determined by specific ELISA and normalized to total Akt content, also determined by ELISA. Values shown are mean ± SEM. *P<0.05, RES vs. SEN groups. Lysate volumes from insulin-stimulated muscle biopsies (4 SEN, 3 RES subjects) containing equivalent amounts of IRS-1 protein (as determined by ELISA) were immunoprecipitated with an antibody against IRS-1. Following SDS-PAGE, Western blotting was performed with a phospho-specific antibody against IRS-1 Ser³¹² (E), or with an anti-phosphotyrosine antibody (F). Blots were quantified by scanning densitometry and results expressed as mean ± SEM. *P<0.0005, RES vs. SEN groups.

Figure 2. Intramyocellular lipid content and body fat stores are increased in insulin resistant, non-obese subjects.

IMCL content of soleus muscles determined by 3D ¹H-MRS multiple voxel technique and expressed as percent H₂O peak (A). Values for percent body fat as determined by DXA are plotted separately males and females to show the independence of gender on the relationship between adipose stores and insulin action (B). Insulin sensitive (black bars) and resistant (white bars) groups are segregated based on results from hyperinsulinemic, euglycemic clamp as described. Values shown are mean ± SEM. *P<0.05, RES vs. SEN groups.