Comparison of the phosphorylation of human erythrocyte spectrin in the intact red cell and in various cell-free systems.

Knowledge of the position and characteristics of the four spectrin phosphorylation sites (Harris, H. W., Jr., and Lux, S. E. (1980) J. Biol. Chem. 255, 11512-11520) was used to study the kinetics of spectrin phosphorylation in the intact red cell. Incubation of intact erythrocytes in the presence of [32P]orthophoshate produced a simultaneous increase in the specific activities of all the spectrin phosphorylation sites. The dephosphorylation of spectrin followed an identical pattern. Spectrin phosphate turnover in intact red cells was then quantitatively compared to isotope labeling patterns produced by phosphorylation of spectrin in various cell-free systems ("in vitro phosphorylation"). 33P-Labeled spectrin dimer, prepared by preincubation of red cells in [33P]orthophosphate, was phosphorylated by a spectrin kinase preparation (Hosey, M. M., and Tao, M. (1977) Biochim. Biophys. Acta 482, 348-357) and [gamma-32P]ATP. Under conditions where no dephosphorylation occurred as judged by loss of 33P label, 0.46 +/- 0.1 mol of 32P-labeled phosphate was incorporated/mol of spectrin dimer. The exogenous 32P label was located in a position and ratio identical with that of the endogenous 33P lebel. Similar results were obtained when the membrane-bound spectrin present in ghosts was phosphorylated using the procedure of Birchmeier and Singer (1977) J. Cell Biol. 73, 647-659). These data indicate that within the intact red cell all of the spectrin phosphates possess an identical rate of isotope exchange and approximately 90% of the spectrin phosphorylation sites are occupied. The remaining 10% may be phosphorylated in vitro by soluble or membrane-bound spectrin kinase in a pattern that is identical with that of the intact erythrocyte.