Kinetic Evidence for an Activation Step Following Binding of Human Interferonα2 to the Membrane Receptors of Daudi Cells

Abstract

A single species of human interferon .alpha. (IFN.alpha.) was labeled with 125I to high incorporation for binding studies on the human B-lymphoblastoid cell line, Daudi, whose growth is inhibited by low doses of IFN, the effect being saturated at .apprx. 100 U/ml (25 pM). The radiolabeled IFN was shown to be fully active and the binding affinity to cellular sites was shown to be unchanged by iodination. Experimental conditions were standardized such that binding and cell growth experiments could be performed on the same initial culture of cells. 125I-labeled IFN.alpha.2 (IFN.alpha. prepared from Escherichia coli carrying human .alpha.2 gene) was added to exponentially growing cultures (mean specific growth rate 0.77 .+-. 0.07 days-1) at a mean concentration of 235,000 .+-. 20,000 cells/ml. Two types of binding could be discerned on growing cultures: the first with a transient peak followed by a loss or discharge of available sites, the second reaching equilibrium some 3 h after the addition of IFN. Large differences in the apparent dissociation constants were evident. The affinity of binding at the steady-state, appeared to be much higher. An analysis of the displacement rates for bound IFN suggested that the 2 reactions were occurring consecutively over the whole of the dose range studied. In this dose range Daudi cells would eventually stop growing at all doses and the rates of deceleration of cellular growth were linearly proportional to the dose of IFN in a double-reciprocal plot (i.e., in analogy to Michaelis-Menten kinetics). A good congruence was found between the equilibrium constants for binding and for growth inhibition. The amount of IFN bound at steady state thus determines the rate at which growth is inhibited. Evidently, the first reaction represents binding of IFN to surface receptors, and the second transfer of IFN to an activation complex on the cell membrane. Appropriate models and their general applicability to IFN action are discussed.