Enzymes work by solvation substitution rather than by desolvation.

Considerable attention has recently been drawn to the hypothesis that enzymes catalyze their reactions by displacing solvent and creating an environment similar to the gas phase for the reacting substrates. This "desolvation hypothesis" is reexamined in this paper by defining a common reference energy for reactions in various environments. It is argued that consistent attempts to describe the actual energetics of enzymatic reactions, taking either gas phase or solution as a reference, would contradict the above hypothesis. That is, the enzyme does remove water molecules from its substrate, but substitutes these molecules for another polar environment (namely, its active site). By taking amide hydrolysis as an example, we use experimentally estimated solvation energies and analyze the reaction profile in the gas phase, in solution, and in enzyme active sites. We show that the gas-phase reaction is characterized by an enormous activation barrier (associated with forming the charged nucleophile from neutral fragments), although the nucleophilic attack is essentially barrierless. On the other hand, the enzyme and solution reactions are found to have similar reaction profiles, with a lower activation barrier for the enzymatic reaction. Presumably, the fact that previous analyses of this problem did not involve the construction of the relevant thermodynamic cycles (and quantitative estimates of the corresponding solvation energies) led to the desolvation hypothesis. Our conclusion is that enzyme active sites provide specific polar environments that do not resemble the gas phase but that are designed for electrostatic stabilization of ionic transition states and that "solvate" these states more than water does.