Cooperative interactions in the binding of allosteric effectors to phosphoenolpyruvate carboxylase.

Phosphoenolpyruvate carboxylase from Salmonella typhimurium is shown to consist of four subunits of molecular weight about 100,000. Guanidinium chloride and 0.1% sodium dodecyl sulfate only partially dissociate the carboxymethylated subunits into monomers of molecular weight of about 50,000. Only one NH*-terminal residue, methionine, is found by the dansylation procedure, which together with peptide mapping results suggests that the enzyme is made up of identical subunits. The enzyme binds 4 moles of L-aspartate and 4 moles of acetyl coenzyme A, i.e. each subunit, although it seems to be made up of two monomers, carries only one binding site for each one of these ligands. The binding of aspartate in the absence of allosteric activators (acetyl-CoA, guanosine triphosphate, and fructose diphosphate), or with the activators present singly, is hyperbolic. The dissociation constant of aspartate is 0.09 mM, but this value varies between 0.1 mM and 0.25 mM in the presence of the various activators. When two activators are present simultaneously, not only does the binding curve for aspartate become sigmoidal but the dissociation constant increases drastically. The dissociation constant for aspartate, for instance, in the presence of acetylCoA and fructose diphosphate together is about 1.5 mM. The binding of acetyl-CoA to the enzyme, in contrast to aspartate, is sigmoidal at low concentrations, but the binding curve shows “tailing” at higher concentration of this activator. A likely cause of this tailing is that acetyl-CoA binds nonspecifically at higher concentrations to some other site on the enzyme surface. In the presence of a second activator, such as GTP, the binding curve for acetyl-CoA becomes hyperbolic and the dissociation constant for this ligand is considerably reduced. Neither magnesium nor manganese is required for the binding of the various ligands on the enzyme surface. The equilibrium binding results are interpreted on the basis of a two-state model, where the enzyme in the absence of any ligands is assumed to be present predominantly in a state which binds the inhibitor, aspartate, preferentially. The cooperative interaction of activators re-