Is allostery an intrinsic property of all dynamic proteins?

Allostery involves coupling of conformational changes between two widely separated binding sites. The common view holds that allosteric proteins are symmetric oligomers, with each subunit existing in "at least" two conformational states with a different affinity for ligands. Recent observations such as the allosteric behavior of myoglobin, a classical example of a nonallosteric protein, call into question the existing allosteric dogma. Here we argue that all (nonfibrous) proteins are potentially allosteric. Allostery is a consequence of re-distributions of protein conformational ensembles. In a nonallosteric protein, the binding site shape may not show a concerted second-site change and enzyme kinetics may not reflect an allosteric transition. Nevertheless, appropriate ligands, point mutations, or external conditions may facilitate a population shift, leading a presumably nonallosteric protein to behave allosterically. In principle, practically any potential drug binding to the protein surface can alter the conformational redistribution. The question is its effectiveness in the redistribution of the ensemble, affecting the protein binding sites and its function. Here, we review experimental observations validating this view of protein allostery.