The entrapment of therapeutic agents in resealed erythrocyte 'ghosts' and their fate in vivo.

Zones were cut from gels run in 0.1 % Triton X-100 put in the ‘starting’ position on the surfaces of fresh gels and refocused. In every case, the proteins moved to positions equivalent to those they had originally occupied in the first gel. The band patterns of plasma membranes of hepatoma D23 and the normal rat hepatocyte were highly characteristic, with many bands common to preparations from the two tissues (Fig. 1). The most basic zones had an apparent PI of 9.5 and the most acidic of 2.9. Most of the periodate-Schiff-positive bands lay between pH 4.5 and 2.9 and in this region most of the zones stained by Coomassie Blue also stained with periodate-Schiff reagents. Rat liver mitochondria gave a markedly different pattern of zones compared with either of the plasma-membrane preparations, and only a few zones were common to both types of membrane. Accordingly the mitochondria1 membranes were chosen as suitable test pieces for mixing with the plasma membranes to determine whether new, mixed ‘hybrid’ bands were produced, such as might arise from micelle formation. No such bands were observed. It was concluded that isoelectric focusing in non-ionic detergents offered a valuable means of comparing membranes and that none of the differences observed between the D23 hepatoma and the normal hepatocyte could be ascribed to the formation of micellar artifacts.