PROTEIN AFFINITIES FOR SMALL MOLECULES - CONCEPTIONS AND MISCONCEPTIONS

There is much confusion and error in published treatments of data for multiple binding of ligands (e.g., substrates) by proteins (e.g., enzymes). There is a widespread impression that if the equilibrium binding, r, of ligand, A, by a protein with n sites can be fitted to an equation with two hyperbolic terms, i.e., r= nαkα(A) 1+kα(A)+ nβkβ(A) 1+kβ(A) (nα+nβ=n) then kβ and kβ are always the intrinsic binding constants for two sets of sites. Such a conclusion is often incorrect. For example, in many cases, the protein is constituted of identical protomers with initially identical sites for binding ligands, and yet graphical representations of the binding data appear to behave as if the sites are partitioned between two classes. Although the use of a linear combination of hyperbolic terms to represent binding of ligands by macromolecules always yields empirical parameters kα, kβ ... kλ, they cannot correspond to site binding constants when there are interactions between sites. In some circumstances their values may even be imaginary, complex numbers. On the other hand, stoichiometric binding constants can be assigned unambiguously without making any assumption regarding the nature of the interactions among binding sites. These principles are illustrated concretely by analyses of binding measurements for several different proteins containing two to six sites. © 1979.