Human epidermal growth factor (EGF) and human transforming growth factor alpha (TGF-alpha) are structurally related polypeptide growth factors that exert their mitogenic activity through interaction with a common cell-surface receptor, the epidermal growth factor receptor (EGFR). The biological effect induced by these two ligands is quantitatively similar in most cases; in some test systems, however, TGF-alpha functions as a more potent form of EGF. In this study, we have compared EGF, TGF-alpha and ten previously described chimaeras of these two ligands in terms of their ability to generate a mitogenic response in cells carrying the human EGFR, and observed that three of the mutant growth factors (E3T, E4T and T3E4T) are mitogenic at concentrations 10-fold lower than that of either wild-type EGF or TGF-alpha. No difference in tyrosine kinase activity of the receptor towards an external substrate was observed after binding of the various mutants. It has been established before [Ebner and Derynck (1991) Cell Regulation 2, 599-612] that EGF and TGF-alpha differ in the processing of the receptor-ligand complex after internalization, as a result of their different pH sensitivities of receptor binding. Similar measurements on our chimaeric mutants revealed that the above superagonists show an enhanced pH dependence of binding in comparison with EGF. Furthermore, induction of receptor recycling by these superagonists is largely comparable with that induced by TGF-alpha. No superagonistic behaviour was observed on a cell-line containing an EGFR/erbB-2 chimaera which does not show ligand-induced internalization. These data show that EGF/TGFalpha chimaeras can be more active than the naturally occurring ligands, and that receptor recycling after ligand-induced internalization seems to be a prerequisite for this phenomenon.

Quantitative analysis of a ligand-protein interaction relates binding to the free concentration of ligand molecules in solution. A theoretical analysis is presented herein, by which intermolecular interactions can be described as a function of the added concentrations of ligand molecules. Following this analysis, ligand binding displacement and saturation curves can be converted directly into a linear form, even when nonradioactively labeled ligands are used to detect the ligand-protein interaction. From the linearities obtained, relevant binding parameters, including the binding dissociation constant, can be calculated. On the basis of this analysis, binding parameters have been characterized for the interaction between biotin-protein A and immunoglobulins, using ELISA-type detection, and for the interaction of a fluorescently labeled fatty acid with a specific fatty acid binding protein.

No abstract

No abstract

No abstract