Adipocyte intrinsic membrane proteins were solubilized in sodium cholate containing buffer, added to cholate-dispersed phospholipids, and reconstituted on Sephadex G-50 columns as previously described [Shanahan, M. F., & Czech, M. P. (1977) J. Biol. Chem. 252, 8341-343], Reconstituted vesicles (0.02-μm diameter) which eluted from the columns were subjected to a freeze-thaw step which resulted in extensive vesicle fusion and enlargement. Brief sonication of the latter preparation reduced the size of the vesicles (0.1-0.5-μm diameter) which then exhibited a time-dependent uptake of D-[3H]glucose that was inhibited by 50 μm cytochalasin B or 90 mM D-glucose and 3-O-methylglucose but not by L-glucose. Cytochalasin B sensitive D-glucose transport activity was obtained in reconstituted vesicles irrespective of the monovalent cation (KCl, NaCl, or choline chloride) used in their preparation and incubation. Transport activity was absolutely dependent upon the presence of divalent cations. Both Ca2+ and Mg2+ were active in this role and had to be added before the freeze-thaw and sonication steps to be effective. Cholate-solubilized intrinsic membrane proteins were fractionated prior to reconstitution by a number of chromatographic techniques in order to further characterize and purify the hexose transporter. Sepharose 6B chromatography in the presence of 0.5% sodium cholate indicated that the transport protein has a Stokes radius between 60 and 80 Å. Hydroxylapatite chromatography in cholate-containing phosphate buffer resolved the solubilized protein into a void peak and two broad peaks. The proteins in the void peak and in the peak eluting between 15 and 100 mM K2HPO4-KH2PO4 contained up to 80 of the total protein, yet exhibited no transport activity when reconstituted into artificial liposomes. The remaining proteins of the second peak which eluted between 150 and 350 mM K2HPO4-K2HPO4 contained all the D-glucose transport activity. No unique protein bands were visible following dodecyl sulfate gel electrophoresis of the transporting fraction compared to the nontransporting fractions, suggesting that the transport protein is a very minor component of the adipocyte membrane. in concert with this conclusion, passage of the intrinsic membrane proteins through a column of concanavalin A—agarose completely removed the two major adipocyte intrinsic membrane glycoproteins (94000 and 78000 daltons) as well as several minor glycoproteins without depleting the transport activity. Adsorption by five other immobilized lectins also failed to prevent subsequent reconstitution of D-glucose transport activity into phospholipid vesicles. All three major protein bands of the cholate-solubilized extracted membrane (94 000, 78 000, and 68 000 daltons) and the multiple minor protein bands could be resolved from the glucose transporter by using concanavalin A-agarose and Sepharose 6B chromatography sequentially. This resulted in substantial purification of the transport system. It is concluded that (1) the adipocyte D-glucose transporter is apparently devoid of oligosaccharide units that can bind to concanavalin A, a lectin known to activate hexose transport in intact fat cells, and (2) the D-glucose transporter is only a minor component of the adipocyte intrinsic membrane protein and exhibits a Stokes radius of 60-80 Å. This size exceeds the width of the plasma membrane phospholipid matrix, consistent with a transmembrane disposition. © 1980, American Chemical Society. All rights reserved.