? Reduction in the expression of the glucose transporter-4 in adipose tissue is one of the earliest changes that occur in human conditions characterized by insulin resistance, including type 2 diabetes. Mice harboring an adipose tissue-specific genetic knockout of Glut4 display reduced insulin-stimulated glucose transport in adipose tissue and develop systemic insulin resistance affecting hepatic glucose production and skeletal muscle glucose transport. Since only adipose tissue is genetically altered in this model, we hypothesized that one or more circulating factors causing systemic insulin resistance may be secreted into the circulation by adipose tissue with reduced Glut4 expression. We used Affymetrix global gene analysis to identify retinol binding protein-4 (RBP4) as a candidate molecule for mediating systemic insulin resistance in this model. The mRNA of RBP4 is up-regulated in adipose tissue, but not in liver, of adipose-Glut4 knockout mice. Serum levels of RBP4 are increased in adipose-Glut4 knockout mice and five other mouse models of insulin resistance, suggesting that RBP4 may be a marker for insulin resistance or hyperinsulinemia. However, we find that RBP4 knockout mice exhibit a phenotype of increased sensitivity to insulin. Thus, elevations of circulating RBP4 may serve not only as a marker but as an actual cause of systemic insulin resistance. In humans we find a strong correlation between serum RBP4 levels and insulin resistance measured by euglycemic hyperinsulinemic clamp, indicating that RBP4 could be a novel endocrine mediator of systemic insulin resistance in humans. To test the hypothesis that RBP4 is a cause of insulin resistance, we will analyze the metabolic features of mice with genetic alterations in RBP4 expression. We will test the effects of pharmacologically altering circulating RBP4 levels, either by treating with fenretinide, a drug that causes urinary wasting of RBP4, or by injecting purified RBP4. We will investigate potential mechanisms, both retinoid-dependent and retinoid-independent, by which RBP4 may influence insulin-glucose homeostasis ? ?