Analysis of the colloid osmotic step of complement-mediated immune hemolysis.

Abstract

After the binding of C9 to EAC1-8, a number of further reactions are required before cell lysis occurs and hemoglobin is released. The final step in this reaction sequence is the colloid osmotic step that occurs after the formation of a red cell intermediate (EAC1-9doomed) with a channel across its membrane. Whether an EAC1-9doomed cell will release its hemoglobin depends on the functional size of the transmembrane channel or channels and the size of the solute in which the cells are suspended. At low C9 to SAC1-8 ratios, the number of EAC1-9doomed cells that will release their hemoglobin decreases as the size of the solute in which the cells are suspended increases.Since SAC1-8 is multivalent for C9, different SAC1-9 complexes containing different numbers of C9 molecules can be formed depending on the C9 concentration used to prepare EAC1-9doomed. If the dimensions of a lytic channel are associated with the number of C9s in the complex, then differences in the extent of lysis from EAC1-9doomed prepared at varying SAC1-8:C9 ratios would be expected in solutes of differing molecular size. Although this explanation would account qualitatively for the data, a least squares fit of a mathematical formulation of the hypothesis shows severe deviation between the experimental and calculated results.The nature of the deviation between the calculated and experimental results suggested the possibility that the binding of C9 to SAC1-8 is negatively cooperative. We therefore extended the model to allow interaction between bound C9 molecules. In this model, an adjustable interaction parameter b is introduced such that b < 1 implies negative cooperativity and b > 1 implies positive cooperativity. This extended model provided a significantly better fit (99% confidence) to sets of data for which end point lysis was low (<35%; i.e., most erythrocytes had either 0 or 1 SAC1-8 complex). The value of b associated with the best fit was less than 1, implying that binding is negatively cooperative.The extended model was tested further against data for which end point lysis was high (>75%; i.e., a substantial portion of erythrocytes have more than 1 SAC18). Deviations between calculated and experimental points were severe, suggesting the possibility that SAC18 complexes do not act independently. A modified model, allowing SAC1-8 interactions, but with no additional parameters, gave an excellent fit to the data, and was significantly better (99.9% confidence) than the fit without aggregation. The average value of the equilibrium constant for C9 interacting with C8 sites in the absence of cooperativity (low degrees of occupancy) was found to be (4.3 ± 0.7) × 1011 M−1.