Why do crown ethers activate enzymes in organic solvents?

One of the major drawbacks of enzymes in nonaqueous solvents is that their activity is often dramatically low compared to that in water. This limitation can be largely overcome by crown ether treatment of enzymes. In this paper, we describe a number of carefully designed new experiments that have improved the insights into the mechanisms that are operative in the crown ether activation of enzymes in organic solvents. The enhancement of enzyme activity upon addition of 18-crown-6 to the organic solvent can be reconciled with a mechanism in which macrocyclic interactions of 18-crown-6 with the enzyme play an important role. Macrocyclic interactions (e.g., complexation with lysine ammonium groups of the enzyme) can lead to a reduced formation of inter- and intramolecular salt bridges and, consequently, to lowering of the kinetic conformational barriers, enabling the enzyme to refold into thermodynamically stable, catalytically (more) active conformations. This assumption is supported by the observation that the crown-ether-enhanced enzyme activity is retained after removal of the crown by washing with a dry organic solvent. A much stronger crown ether activation is observed when 18-crown-6 is added prior to lyophilization, and this can be explained by a combination of two effects: the before-mentioned macrocyclic complexation effect, and a less specific, nonmacrocyclic, lyoprotecting effect. The magnitude of the total crown ether effect depends on the polarity and thermodynamic water activity of the solvent, the activation being highest in dry and apolar media, where kinetic conformational barriers are highest. By determination of the specific activity of crown-ether-lyophilized enzyme as a function of the enzyme concentration, the macrocyclic crown ether (linearly dependent on the enzyme concentration) and the nonmacrocyclic lyoprotection effect (not dependent on the enzyme concentration) could be separated. These measurements reveal that the contribution of the nonmacrocyclic effect is significantly larger than the macrocyclic refolding effect