Cross-Linking of Glycolipids in Erythrocyte Ghost Membrane*

Human erythrocyte ghosts were chemically cross-linked with a family of bifunctional cross-linking reagents, imidoesters. Since the reagents are identical in their chemical structures, except for the number of methylene groups, they differ by that distance between the functional groups. The maximum distance is 4.9 A in dimethyl malonimidate, 6.1 A in dimethyl succinimidate, 8.6 A in dimethyl adipimidate, and 11 A in dimethyl suberimidate. Cross-linked membrane glycoproteins and glycolipids were solubilized in 1% sodium dodecyl sulfate and analyzed by electrophoresis on sodium dodecyl sulfate-polyacrylamide gels. After cross-linking, two prominent new bands of cross-linked products, GP-A and GP-B, appeared. Cross-linking of the new bands was dependent on the length of the reagents. GP-A could be produced by all of the reagents, but GP-B only by dimethyl adipimidate (8.6 A) and dimethyl suberimidate (11 A). In our previous report, GP-A was shown to be a complex of two species of the glycoproteins. In this study, it is shown that the glycolipids and proteins are components of GP-B. The conclusion is based on the following observations. (a) The GP-B formation was dependent on the distance between the functional groups of the reagents. (b) GP-B could not be produced by reacting the ghosts with methyl butyroimidate, a monofunctional analogue of dimethyl adipimidate. (c) GP-B could be stained by periodate-Schiff’s reagent and by Coomassie brilliant blue. (d) Cleavage of the crosslinks in GP-B by ammonolysis produced three protein bands and a lipid band which was positive to periodate-Schiff’s reagent. (e) The cleaved protein bands could be stained by Coomassie brilliant blue and not by periodate-Schiff’s reagent. (f) Synthetic aminophospholipids and trilinolein could not be stained by periodate-Schiff’s reagent. (g) When ghosts were prepared from erythrocytes which were labeled with tritium (catalyzed by galactose oxidase), GP-B also appeared to be labeled. (h) Upon cleavage of labeled GP-B, most of the label was found in the lipids and not in the proteins. (i) GP-B was not labeled when erythrocytes were labeled without galactose oxidase treatment. The cross-links of the glycolipids and the proteins suggest possible specific association between them. The following observations substantiate the premise. (a) The glycolipids and the proteins always exhibited an identical electrophoretic mobility, regardless of the acrylaxnide concentration in the