Affinity-, Partition- and Permeability Properties of the Human Red Blood Cell Membrane and Biomembrane Models, with Emphasis on the GLUT1 Glucose Transporter

Abstract: The human glucose transporter GLUT1 is abundant in red blood cells, the blood-brain barrier and epithelial cells, where it mediates the transport of the energy metabolite, glucose. In the present work some properties of GLUT1, including affinity binding of both substrates and inhibitors, transport rates as well as permeabilities of aromatic amino acids and drug-membrane interactions were analyzed by chromatographic methods.Reconstitution by size-exclusion chromatography on Superdex 75 from a detergent with a low CMC that provides monomeric GLUT1 was examined regarding D-glucose- and CB binding as well as D-glucose transport. Upon steric immobilization in Superdex 200 gel beads, residual detergent could be washed away and dissociation constants in the same range as reported for binding to GLUT1 reconstituted from other detergents were obtained. The transport rate into the GLUT1 proteoliposomes was low, probably due to residual detergent. Binding to GLUT1 at different pH was analyzed and the affinity of glucose and GLUT1 inhibitors was found to decrease with increasing pH (5–8.7). The average number of cytochalasin B-binding sites per GLUT1 monomers was, in most cases, approximately 0.4. GLUT1 may work as a functional monomer, dimer or oligomer. To determine whether GLUT1 was responsible for the transport of the aromatic amino acids tyrosine and tryptophan, uptake values and permeabilities of these amino acids into liposomes and GLUT1 proteoliposomes were compared to the permeabilities of D- and L- glucose in the same systems. Dihydrocytochalasin B was identified to be a new inhibitor of tyrosine and tryptophan transport into red blood cells. Ethanol turned out to inhibit the specific binding between CB and GLUT1 and also to decrease the partitioning of CB and drugs into lipid bilayers. A capacity factor for drug partitioning into membranes that allows comparison between columns with different amount of immobilized lipids was validated, and turned out to be independent of flow rate, amount of lipids and drug concentration in the ranges tested.