Fructose consumption has increased in modern America, and fructose consumption may play a role in metabolic diseases including obesity and diabetes. Fructose transport is regulated by a protein family of glucose transporters (GLUTs), and the fructose transporter GLUT5 is expressed in enterocytes from the small intestine. Thioredoxin-Interacting Protein (Txnip) is a member of a mammalian protein superfamily that contains arrestin-like domains. Here, we present new preliminary data showing that Txnip interacts with the fructose transporter GLUT5. We also show that overexpression of Txnip increases cellular fructose uptake in vitro, while deletion of Txnip inhibits fructose uptake. Our preliminary in vivo experiments further demonstrate that a high fructose diet causes non-alcoholic fatty liver disease in wild type mice, while mice deficient in Txnip have reduced intestinal fructose transport and hepatic fat deposition following a fructose diet. These data support a new overall theory that Txnip enhances fructose transport through specific molecular interactions with GLUT5. The experiments proposed here will likely reveal the molecular basis for a new pathway in fructose metabolism. Our Specific Aims are: Specific Aim 1 will test the hypothesis that a specific molecular interaction through an alpha arrestin-domain in Txnip regulates GLUT5 and GLUT2 functions. Here we will dissect the molecular specificity of this interaction by structure-function analysis. Specific Aim 2 will test the hypothesis that enterocyte Txnip regulates intestinal fructose transport and promotes hepatic steatosis in vivo. Targeted loss of function of Txnip will be used to test this hypothesis in vivo. We will generate intestine-specific Cre-mediated deletion of Txnip in mice, by crossing the Txnip flox/flox mouse that we engineered with a transgenic mouse bearing a Cre recombinase expressed under the control of the villin promoter. Specific Aim 3 will test the hypothesis that Txnip increases intestinal fructose transport through a redox-mediated mechanism in vivo. In this aim, we will test if GLUT5 regulation is controlled by Txnip through a thioredoxin-dependent redox-mediated mechanism, using a novel mouse with a “knock- in” mutation of Txnip that we have just generated. This mouse has a single amino acid change that eliminates binding of Txnip to thioredoxin, and thus renders Txnip function independent of thioredoxin redox state.