Partition in two-polymer aqueous phases reflects differences between membrane surface properties of erythrocytes, ghosts and membrane vesicles.

Partition of cells (defined as the percentage of added cells in the top phase) in dextran-polyethylene glycol aqueous two-phase systems is an extremely sensitive indicator of cell surface properties. By appropriate choice of polymer concentrations and salt composition and concentration one can, to some extent, select the membrane properties (e.g. charge (Walter, H. (1975) in Methods in Cell Biology (Prescott, D.M., ed.), Vol. 9, pp. 25-50, Academic Press, New York), lipid composition (Walter, H, Krob, E.J. and Brooks, D.E. (1976) Biochemistry, 15, 2959-2964)) that determine the partition behavior of cells (or membranes). In the present experiments we have obtained by partition information on separability, alteration and ensuing heterogeneity during preparation of human erythrocytes, ghosts and rightside-out and inside-out vesicles from human erythrocyte membranes. A phase system in which partition is greatly (although not exclusively) dtermined by membrane charge was used. It was found that: 1. The partition coefficients (and hence the membrane surfaces) of ghosts and of rightside-out vesicles differ from those of the erythrocytes from which they are derived. 2. Rightside-out and inside-out vesicles have different countercurrent distribution patterns with the rightside-out vesicles having the higher partition coefficient (probably reflecting higher membrane charge of the latter). 3. Rightside-out vesicles are highly heterogeneous membrane populations as evidenced by broader than theoretical countercurrent distribution curve which, in some experiments, even split into two populations outright. This heterogeneity is particularly interesting since the enzyme markers used to establish vesicle sidedness (Steck, T.L. (1974) in Methods in Membrane Biology (Korn, E.D., ed), Vol. 2, pp. 245-281, Plenum Press, New York) indicate "pure" rightside-out preparations. 4. Inside-out vesicles could, under no conditions tested, be rendered totally free of rightside-out vesicles according to the enzyme markers used to indicate vesicle sidedness (Steck, T.L. (1974) in Methods in Membrane Biology (Korn, E.D., ed), Vol. 2, pp. 245-281, Plenum Press, New York). A discussion of factors involved in the partition of the vesicles and ghosts together with analogies to the partition behavior of stored or in vitro modified human red blood cells is presented.