Increased skeletal muscle tumor necrosis factor-alpha and impaired insulin signaling persist in obese women with gestational diabetes mellitus 1 year postpartum

Abstract

Objective: Women with gestational diabetes mellitus (GDM) demonstrate chronic and progressive insulin resistance and a markedly increased risk of converting to type 2 diabetes after pregnancy. However, the cellular mechanisms underlying this insulin resistance are unknown.

Research design and methods: We investigated the progression of insulin resistance in nine obese women with GDM during late pregnancy (30-36 weeks) and 1 year postpartum. Skeletal muscle biopsies were obtained at each visit, and insulin resistance was determined by the hyperinsulinemic-euglycemic clamp technique.

Results: Insulin resistance was not significantly improved in GDM women (4.1 +/- 0.4 vs. 5.8 +/- 1.1 10(-2) mg x kg FFM x min(-1)/microU x ml(-1)). Subjects did not experience significant weight loss postpartum. Body weight, fat mass, fasting glucose, and plasma tumor necrosis factor (TNF)-alpha remained higher 1 year postpartum than seen in previously studied normal glucose-tolerant women. Skeletal muscle TNF-alpha mRNA was elevated five- to sixfold in GDM women and remained higher 1 year postpartum. While levels of insulin receptor (IR), IR substrate (IRS)-1, and p85 alpha improved postpartum, insulin-stimulated IR tyrosine phosphorylation and receptor tyrosine kinase activity did not significantly improve postpartum in GDM. The levels of (312)Ser-IRS-1 also did not improve postpartum and correlated with TNF-alpha mRNA (r(2) = 0.19, P < 0.03), consistent with a state of subclinical inflammation and chronic skeletal muscle insulin resistance.

Conclusions: These results suggest the mechanisms underlying chronic insulin resistance in GDM women may be driven by increased inflammation that impinges on the IR and IRS-1 signaling cascade in skeletal muscle. These findings have important implications for the health of GDM women during subsequent pregnancies and their risk for progression to type 2 diabetes.

Figures

FIG. 1

Insulin sensitivity measured during hyperinsulinemic-euglycemic clamps performed during late pregnancy (30–36 weeks) and repeated ~1 year postpartum. Data are means ± SE, n = 9. No significant difference was found (P = 0.1034). The M value was calculated for the final 150–180 min of the clamp. Units are expressed relative to fat-free mass (FFM).

FIG. 2

Body weight based on self-reported weights prepregnancy (□) and measured weights during late pregnancy (30–36 weeks; ) and ~1 year postpartum (■). Data are means ± SE, n = 9. *P < 0.0001 vs. prepregnancy, **P < 0.0001 vs. late pregnancy, #P < 0.005 vs. prepregnancy, and ##P < 0.05 vs. prepregnancy. Data for NGT patients were reported previously (14) and are included for reference purposes.

FIG. 3

Basal and insulin-stimulated IR tyrosine phosphorylation (A) and IRTK activity (B) measured in isolated insulin receptors from skeletal muscle obtained during late pregnancy and ~1 year postpartum. The IR concentration was measured in each biopsy based on a standard curve with purified insulin receptors, and an equal concentration of IR protein was tested for IR tyrosine phosphorylation and IRTK activity by ELISA as outlined previously (14). Data are means ± SE for obese women with GDM (n = 5). *P < 0.001 vs. basal (no insulin).

FIG. 4

Skeletal muscle insulin signaling proteins from vastus lateralis obtained during late pregnancy and ~1 year postpartum in obese GDM subjects. Equivalent amounts of protein isolated from skeletal muscle biopsies were subjected to SDS-PAGE and blotted with respective antibodies. Values are means ± SE (n = 9) and expressed as percent change from late pregnancy to postpartum. *P < 0.05 vs. late pregnancy, **P < 0.005 vs. late pregnancy.

FIG. 5

TNF-α during late pregnancy (30–36 weeks; □) and ~1 year postpartum (■). Data are presented as mean ± SE for obese women with GDM during pregnancy (n = 9) and women with NGT during pregnancy (n = 9). A: Plasma TNF-α was measured by ELISA. Data are mean ± SE, n = 9. *P < 0.05. B: Skeletal muscle TNF-α mRNA expression was measured by quantitative real-time PCR. Data are mean ± SE, n = 9. *P < 0.005 vs. NGT during late pregnancy, **P < 0.02 vs. GDM during late pregnancy. Data for NGT patients were reported previously (14) and are included for reference purposes.

FIG. 6

A: Correlation between the change in 312Ser-IRS-1 and the change in ISI from late pregnancy to postpartum in NGT and GDM subjects. B: Correlation between 312Ser-IRS-1/total IRS-1 and TNF-α mRNA in NGT and GDM subjects.