Insulin-like growth factor-binding protein-3 (IGFBP-3) interacts with the type II nuclear receptors retinoid X receptor (RXR)alpha and retinoic acid receptor-alpha and modulates their transcriptional activity. Peroxisome proliferator-activated receptor (PPAR)gamma, a related nuclear receptor that dimerizes with RXRalpha, plays an important role in adipocyte differentiation. IGFBP-3 is regulated during adipocyte differentiation, but its role in this process is unknown. We demonstrate that IGFBP-3 interferes with the PPARgamma-dependent processes of adipocyte differentiation and maintenance of the gene expression characteristic of mature adipocytes. Treatment of adipocytes with exogenous IGFBP-3, but not an IGFBP-3 mutant that does not bind RXRalpha or PPARgamma, decreased markers of adipocyte differentiation, PPARgamma, and resistin but increased the preadipocyte marker plasminogen activator inhibitor-1. Furthermore, expression of human IGFBP-3, but not the IGFBP-3 mutant, by preadipocytes inhibited preadipocyte differentiation as determined by gene markers and lipid accumulation. IGFBP-3 interacted with PPARgamma in vitro and in 3T3-L1 adipocyte lysates and inhibited PPARgamma heterodimerization with RXRalpha in vitro. Wild-type IGFBP-3, but not mutant IGFBP-3, blocked ligand-induced transactivation of PPAR response element in 3T3-L1 cells. The observation that IGFBP-3 inhibits adipocyte differentiation and impacts on the PPARgamma system suggests a role for IGFBP-3 in the pathogenesis of obesity and insulin resistance.

IGFBP-3 may contribute to insulin resistance in adipocytes.

Iodine deficiency in the postpartum period has the potential to affect neonatal neuropsychointellectual development. We performed a cross-sectional study involving 50 postpartum women and their neonates, measuring maternal urine iodine, breast milk iodine, and neonatal thyroid stimulating hormone (TSH), and examining their interrelationships. Women were studied at a median (range) of 4 (3-9) days postpartum. Moderate to severe iodine deficiency (defined by urine iodine concentration < 50 microg/L) was found in 29 of the 50 subjects (58%). The median +/- standard deviation (SD) urine iodine was 46.8 +/- 28.5 microg/L and the mean urine iodine expressed in micrograms per gram of creatinine was 86.6 +/- 45.6. The median (range) breast milk iodine was 84.0 microg/L (25.0-234.0). Breast milk iodine was significantly correlated with urine iodine in micrograms per gram of creatinine (r = 0.52, p < 0.001) but not with urine iodine measured in micrograms per liter (r = 0.19, p = 0.2). Six percent of neonates had whole-blood TSH values of greater than 5 mIU/L. Neonatal TSH levels were positively correlated with higher breast milk iodine (r = 0.42, p = 0.003). There was no significant correlation between neonatal TSH levels and the mother's urine iodine content. There is a high prevalence of iodine deficiency in these lactating postpartum subjects. Urine iodine as micrograms per gram of creatinine is a good predictor of breast milk iodine content. In our study, higher breast milk iodine was correlated with a higher neonatal TSH. The impact of breast milk iodine content on neonatal TSH levels and neuropsychointellectual development needs further study.