Short hydrogen bonds in the catalytic mechanism of serine proteases

Abstract

The survey of crystallographic data from the Protein Data Bank for 37 structures of trypsin and other serine proteases at a resolution of 0.78-1.28 ? revealed the presence of hydrogen bonds in the active site of the enzymes, which are formed between the catalytic histidine and aspartate residues and are on average 2.7 ? long. This is the typical bond length for normal hydrogen bonds. The geometric properties of the hydrogen bonds in the active site indicate that the H atom is not centered between the heteroatoms of the catalytic histidine and aspartate residues in the active site. Taken together, these findings exclude the possibility that short "low-barrier" hydrogen bonds are formed in the ground state structure of the active sites examined in this work. Some time ago, it was suggested by Cleland that the "low-barrier hydrogen bond" hypothesis is operative in the catalytic mechanism of serine proteases, and requires the presence of short hydrogen bonds around 2.4 ? long in the active site, with the H atom centered between the catalytic heteroatoms. The conclusions drawn from this work do not exclude the validity of the "low-barrier hydrogen bond" hypothesis at all, but they merely do not support it in this particular case, with this particular class of enzymes.