Heterogeneity and Unusually High Affinity in the Interactions of Adenosine 3′:5′-Monophosphate with Specific Binding Proteins from Liver and Hepatoma Cells

Abstract

Abstract The binding of adenosine 3':5'-monophosphate (cyclic AMP) to proteins has been further characterized in postmitochondrial supernatant fractions from rat liver and cultured hepatoma (HTC) cells. Possible explanations for the high binding affinity (Kd l 0.5 nm) and nonlinear Scatchard binding plots are considered. Examination of the membrane filtration method used to measure the binding indicates that essentially all of the protein in the extracts is retained and that nonspecific adsorption of cyclic AMP is not a serious problem under the conditions used. At high concentrations of cyclic AMP, the amount of binding increases linearly over a range of protein concentrations. The binding studied is highly selective for cyclic AMP as shown by competition studies with related compounds. The time required to reach binding equilibrium at 0° varies with the pH, the nature of the sample, and the cyclic AMP concentration. The observation of nonlinear Scatchard binding plots requires a wide range of cyclic AMP concentrations, and this in turn necessitates binding times of several hours or more to achieve equilibrium at all points. The kinetics of release of bound cyclic AMP are heterogeneous for both liver and HTC extracts, showing a range of half times from minutes to days. An evaluation of the instability of binding activity under assay conditions shows that this instability cannot cause the upward curvature seen in Scatchard plots, but it could account for the additional plateau or region of downward curvature observed with liver samples. Prior exposure of a liver sample to pH 4 or to elevated temperatures eliminates the component(s) responsible for the plateau and transforms the Scatchard binding curve at pH 4 into one showing only the upward curvature characteristic of HTC samples. The high affinity portion of the heterogeneous Scatchard plots is markedly reduced by increasing the salt concentration in the binding reaction at pH 4, but not at pH 6.5. Treatment with HgCl2 in the cold also produces some selective elimination of the higher affinity portion of the HTC Scatchard binding curve. Although these selective effects on different regions of the HTC Scatchard plot are consistent with the presence of multiple components with differing affinities for cyclic AMP, other explanations cannot be ruled out. Negatively cooperative interactions among identical binding sites would also account for the curvature, and these interactions might be sensitive to some of the agents used. In addition, theoretical considerations show that the interaction between a single binding component and the catalytic subunit of a protein kinase can generate nonlinear Scatchard plots similar to those observed.