The binding of substrates to these enzymes accounts for the selectivity difference between the enzymes:

• The binding site of chymotrypsin (previous page) has a non-polar cleft large enough to accomodate an aromatic ring

• The cleft in trypsin is of similar size to that in chymotrypsin (methylene groups are about 4 A in diameter), but has a negative carboxylate at the bottom, so that it binds residues with positive side chains:

  Binding Pocket of Trypsin

  In the left picture, the active site residues are green; the violet residue is Asp189, at the bottom of the binding cleft, to grab positive side chains

• In elastase, a valine (216) and a threonine (226) replace the glycines at the mouth of the binding pocket, blocking all except small side chains from entering.

  Binding Pocket of Elastase

  The color scheme is the same.

The binding, in cartoon form:

The next step appears to be His57 removing, or strongly H-bonding, a proton from the the Ser195 OH, while the O does a nucleophilic attack on the peptide carbonyl:

The shift of negative charge onto the carbonyl oxygen of the peptide is facilitated by hydrogen bonding of the oxygen to the backbone NH groups of Ser195 and Gly193.

• These form what is called "the oxyanion hole"

At this point we have formed a structure equivalent to the tetrahedral intermediate in the non-enzymatic mechanism.

• The biggest difference is that the two protons that are part of the OH groups are merely hydrogen-bonded in the enzyme

• In the organic mechanism, we write them as fully attached

• This in fact may be an artifact of organic chemists' notation, since they surely are strongly H-bonded in the non-enzymatic case as well

• Some biochemists describe these as "low barrier" hydrogen bonds

Whether a tetrahedral intermediate is formed was a point of considerable contention in the early investigations of protease mechanisms

• Biochemists tend to draw nucleophilic acyl substitutions as if they were S_N2 reactions - all one step.

This ignores a fundamental difference between the two kinds of reaction.

• Nucleophilic substitution, in frontier orbital terms, involves an interaction between the HOMO of the nucleophile and the LUMO of the electrophile (substrate) - a "filled-empty" interaction.

• Alkyl substrates, which undergo S_N2 reactions, have much different LUMOs than acyl substrates, which form tetrahedral intermediates

To illustrate, here's the LUMO of ethyl chloride, CH₃CH₂Cl:

LUMO of Ethyl Chloride

You can see that it involves chiefly the C-Cl bond, and has a considerable back lobe where the nucleophile interacts. Now here is the LUMO of acetyl chloride, CH₃(C=O)Cl:

LUMO of Acetyl Chloride

This LUMO is almost exclusively the C=O p* orbital, meaning that the bond that breaks on nucleophilic attack is the C=O double bond.

• Notice also that the nucleophile must attack perpendicular to the plane of the carbonyl.

Hence there is no way an attacking nucleophile can break the C-Cl bond directly.