A mechanistic study of the spontaneous hydrolysis of glycylserine as the simplest model for protein self-cleavage.

A common feature of several classes of intrinsically reactive proteins with diverse biological functions is that they undergo self-catalyzed reactions initiated by an N→O or N→S acyl shift of a peptide bond adjacent to a serine, threonine, or cysteine residue. In this study, we examine the N→O acyl shift initiated peptide-bond hydrolysis at the serine residue on a model compound, glycylserine (GlySer), by means of DFT and ab initio methods. In the most favorable rate-determining transition state, the serine COO(-) group acts as a general base to accept a proton from the attacking OH function, which results in oxyoxazolidine ring closure. The calculated activation energy (29.4 kcal mol(-1) ) is in excellent agreement with the experimental value, 29.4 kcal mol(-1) , determined by (1) H NMR measurements. A reaction mechanism for the entire process of GlySer dipeptide hydrolysis is also proposed. In the case of proteins, we found that when no other groups that may act as a general base are available, the N→O acyl shift mechanism might instead involve a water-assisted proton transfer from the attacking serine OH group to the amide oxygen. However, the calculated energy barrier for this process is relatively high (33.6 kcal mol(-1) ), thus indicating that in absence of catalytic factors the peptide bond adjacent to serine is no longer a weak point in the protein backbone. An analogous rearrangement involving the amide N-protonated form, rather than the principle zwitterion form of GlySer, was also considered as a model for the previously proposed mechanism of sea-urchin sperm protein, enterokinase, and agrin (SEA) domain autoproteolysis. The calculated activation energy (14.3 kcal mol(-1) ) is significantly lower than the experimental value reported for SEA (≈21 kcal mol(-1) ), but is still in better agreement as compared to earlier theoretical attempts.