Redistribution of Lu gp after microdeformation.

(A) Bright field image of red cell being aspirated into micropipette. (B) Fluorescence micrograph and (C) corresponding intensity profile of Lu gp labeled with fluorescent antibody BRIC 221 in the partially aspirated red cell. The intensity profile indicated that the density of FITC-labeled anti-Lu gp increased markedly at the pipette entrance (ρ_ε) and subsequently decreased toward the aspiration cap (ρ_c). (D) The ratio of entrance to cap density (ρ_ε/ρ_c) of FITC-labeled anti-Lu gp (squares) plotted as a function of aspiration length (L/Rp). For comparison, the upper dashed black line shows prior data of rhodamine phalloidin-labeled actin, indicating the dependence of ρ_ε/ρ_c versus L/Rp for molecules tightly associated with the membrane skeleton. The lower dashed gray line shows prior data of eosin-5-maleimide (EMA)-labeled band 3, and indicates the presence of both skeleton-associated and freely diffusing populations of band 3. The interpretation of the density gradient obtained with the fluorescence-imaged microdeformation technique is based on accumulated data from FIMD experiments on red cell membranes with well-defined membrane protein abnormalities. We know that the steepness of the density gradient of a particular protein is determined by the extent of its linkage to the cytoskeleton by analyzing the density distribution of integral proteins in cells in which the skeletal proteins linking these integral proteins to the cytoskeleton are lacking. Specifically, although glycophorin C in normal cells exhibits a steep gradient, it is freely diffusable in protein 4.1-deficient cells because in these cells it is not linked to the membrane skeleton due to absence of protein 4.1.26,27 Similarly, in ankyrin-deficient cells, the density of band 3 fails to exhibit a gradient40 because ankyrin molecules are not available to link band 3 to spectrin.