PURIFICATION OF STABILIZED BAND 3 PROTEIN OF THE HUMAN ERYTHROCYTE MEMBRANE AND ITS RECONSTITUTION INTO LIPOSOMES

The evidence implicating Band 3 protein of the human erythrocyte membrane as the anion channel is quite substantial. An important part of this proof has been the isolation and reconstitution of this protein into liposomes and subsequent demonstration of its ability to mediate anion transport. Several laboratories have accomplished this; however, the Band 3 preparations in each case have been contaminated with other membrane proteins, notably Bands 4.2, 4.5 and glycophorin (PAS-1). An additional problem has been that isolated Band 3 preparations are notoriously unstable and rapidly form high molecular weight aggregates.;Band 3 was purified from ghosts which were first washed with isotonic saline to remove Band 6 (glyceraldehyde-3-phosphate dehydrogenase) and then extracted overnight with 0.5% Triton. The Triton extract was concentrated and applied to DEAE cellulose column to remove lower molecular weight proteins (mostly in the 4.5 region). Higher salt concentrations were then used to elute a fraction containing mostly Band 3, Band 4.2 and glycophorin and this fraction was then passed through pCMB aminoethyl Agarose 4B affinity gel. During this time glycophorin and some Band 4.2 eluted. The gel was then washed with the same high salt buffer in which the protein was applied and with a low salt buffer before pure Band 3 was eluted with 0.1 mM cysteine. Some additional Band 3 could be eluted with 50 mM cysteine, but this fraction contained some Band 4.2 as well. Upon elution of Band 3, 15 mM mercaptoethanol was added immediately as this was found to keep the protein from aggregating. Preparations routinely containing at least 95% Band 3 ((TURN)1% dimer) and less than 1.5% Band 4.2 have been obtained. Using the above method, pure Band 3 could be isolated from the Triton extract in less than 4 hours.;Band 3, according to sedimentation analysis on sucrose gradients existed as a dimer in buffer from which Triton was removed. After removal of Triton, Band 3 incorporated into liposomes composed of 96% phosphatidyl choline, 4% phosphatidic acid, increased sulfate efflux more than 70-fold over control. When 10 (mu)M 4,4'-diisothiocyano-2,2'-stilbene disulfonate was added to the outside of Band 3-containing liposomes, 30 - 40% of the efflux was inhibited. When the inhibitor was added to the protein before incorporation, nearly complete inhibition was attained. When transport was carried out with no external transportable anions, sulfate efflux was markedly reduced indicating efflux proceeds mainly by an exchange mechanism as in the intact red cell. At 0(DEGREES)C in 36 mM Na(,2)HPO(,4), 15 mM mercaptoethanol, pH 7.50, Band 3 remained functional and monomeric for at least a week.;A new quick method has been developed for the purification of a functional Band 3 protein essentially free from all other contaminants. This method employs the use of pCMB aminoethyl Agarose 4B affinity gel which was synthesized by significant modifications of a method of Cuatrecasas. These modifications ensure complete substitution of the free amino groups on the spacer molecule with pCMB. Using this gel, glycophorin has been removed from Band 3 preparations since this protein does not contain any sulfhydryl groups and does not bind to the pCMB on the gel. This was not possible using the gels synthesized by the method of Cuatrecasas or purchased commercially (Bio Rad, Affi-Gel 501) since these gels absorbed excessive amounts of glycophorin probably due to the binding of this protein to spacer molecules not coupled to pCMB. It should be noted that this modified pCMB affinity gel can also be used to purify glycophorin.