Superposition of Thermolysin and Carboxypeptidase A (Catalytic Sites CPK)
Sequence Alignment (18% Identity)

OK, I believe: we're looking at convergent evolution. Similar function, similar catalytic residues, but almost no sequence homology.

The mechanism, although, of course, it must be a nucleophilic acyl substitution, is not entirely clear. One suggestion, from Hilvert, is below:

Mechanism of the Zn Proteases

• The carbonyl oxygen of the peptide bond to be hydrolyzed is coordinated to the zinc, displacing a water

• The Glu (143 or 270) acts as a general base, facilitating attack of a water on the carbonyl carbon to form the tetrahedral intermediate

• The protein being cleaved is never bound to the enzyme; that is, there is no acyl enzyme intermediate

• In thermolysin, the His231 facilitates the loss of the amine; Lipscomb had suggested that Try248 performed this role in carboxypeptidase A

An alternative is the direct attack of the Glu on the carbonyl carbon:

Alternative Mechanism of the Zn Proteases

• In this case, the acyl enzyme intermediate (actually, an anhydride) is formed, and another step is required to hydrolyze it

• The enzymes can be inactivated by alkylating the Glu

• Since such alkylation severely distorts the enzyme conformation, a direct nucleophilic role for the Glu may be unlikely

A crystal structure from Lipscomb's group provides inferential support for the direct hydrolysis.

Llipscomb fed the carboxypeptidase a transition state inhibitor - a structure that would mimic a tetrahedral intermediate bound to the Zn.

The cystal structure shows the inhibitor with both phosphonate oxygens coordinated to the Zn:

Transition State Inhibitor Bound to Carboxypeptidase (6cpa)

As an exopeptidase, carboxypeptidase has a relatively deep binding site.

• Only the end of the protein chain need fit in.

• Thermolysin is an endopeptidase, like the serine proteases; it has an open, extended cleft for binding substrate.

Carboxypeptidase A (1yme)	Thermolysin (1fjq)