Andrew Shrake

The results of a thermodynamic calculation of the excess heat capacity that is based on experimental observations and that incorporates the effects of ligand binding on the two-state, thermal denaturation of a protein are presented. For a protein with a single-binding site on the native species and at subsaturating concentrations of ligand, bimodal or unimodal thermograms were computed merely by assuming a larger or smaller ligand association constant, respectively. The calculated thermograms for this simplified case show the salient features of those observed by differential scanning calorimetry for defatted human albumin monomer in the absence and presence of three ligands for which the protein has higher, intermediate, and lower affinity (Shrake, A., and Ross, P. D. (1988) J. Biol. Chem. 263, 15392-15399). The computation demonstrates that biphasic unfolding can result from a significant increase in the free energy of denaturation (and the transition temperature) during the course of unfolding due to a substantial increase in free ligand concentration caused by the release of bound ligand by denaturing protein. Such ligand-induced biphasic denaturation does not relate to macromolecular substructure but derives from a perturbation, during unfolding, of the ligand binding equilibrium, which is coupled to the equilibrium between the folded and unfolded protein species. Thus, this bimodality is not limited to thermally induced unfolding but is operative independent of the means used to effect denaturation and therefore must be considered when studying any macromolecular folding/unfolding reaction in the presence of ligand.

Denaturation of defatted human albumin monomer, monitored by differential scanning calorimetry, is monophasic as reflected by the single, resulting endotherm. With low levels of various ligands, biphasic or monophasic unfolding processes are manifested as bimodal or unimodal thermograms, respectively. The greater the affinity of native protein for ligand, the greater is the tendency for biphasic denaturation. We propose that such a biphasic unfolding process arises from a substantial increase in stability (transition temperature) of remaining native protein during denaturation. This increase in stability derives from the free energy of ligand binding becoming more negative due to the release of high affinity ligand by unfolding protein. The tendency for biphasic denaturation is greatest at low (subsaturating) levels of ligand where greatest increases in stability occur. Biphasic unfolding arising from such ligand redistribution results from denaturation of different kinds of protein molecules, ligand-poor and ligand-rich species, and not from sequential unfolding of domains within the same molecule. Differentiating between these two mechanisms is necessary for the correct interpretation of biphasic denaturation data. Furthermore, biphasic unfolding due to ligand redistribution occurs independently of the means used to effect denaturation. The maximum increase in stability due to ligand binding relative to the stability of defatted albumin monomer alone occurs with the intermediate affinity ligand octanoate (22 degrees C) and not with the high affinity ligand hexadecanoate (15 degrees C). This indicates a much greater affinity of denatured albumin for hexadecanoate since increase in stability derives from the difference between free energy of ligand binding to folded and unfolded protein forms.