Proelastase contains a 12 residue segment at the N-terminal end that is cleaved in activation, but retained by a disulfide bond between the N-terminal Cys and Cys125:

Cys-Gly-Asp-Pro-Thr-Tyr-Pro-Pro-Tyr-Val-Thr-Arg

Unfortunately, no crystal structure exists in which these twelve residues are present; all begin with a pair of valines. Here is one of those structures:

Elastase (1lvy)	Catalytic Site and Salt Bridge

Note that the new N-terminal valine (16) forms a salt bridge to Asp194 (red), just like the N-terminal Ile in chymotrypsin.

One may guess that the activation also involves clearing the binding pocket by a similar conformational change.

The similarity between trypsin, chymotrypsin, and elastase is very strong. All three backbones superpose with rmsd= 1.10 A:

Superposition of Chymotrypsin (green), Trypsin (gold) and Elastase (blue)

The catalytic residues are shown as CPK models: Asp102, His57, and Ser195. Clearly the active sites, if not exactly identical, are very, very close. One can just barely see all three colors for these residues.

Here are sequence alignments for chymotrypsin (4cha), trypsin (1tld), and elastase (3est) produced by ClustalW and BioEdit:

Chymotrypsin vs Trypsin vs Elastase

Another example, with another serine protease, to emphasize the point:

Chymotrypsin Aligned with Thrombin
rmsd = 0.96 A; 30% identity, 42% similarity

What we see here is a clear example of divergent evolution.

• All are derived from a common ancestral serine protease, and are described as homologous.

Evolution can converge on functionality also. The first crystal structure of a bacterial serine protease, subtilisin, from B. amyloliquefaciens, shows a thoroughly different construction from the mammalian ones, and essentially no sequence homology:

Subtilisin (B. amyloliquefaciens; 2st1)	Catalytic Site

But the enzymes are functionally identical:

• Subtilisin uses the same three catalytic residues: Asp32, His64, and Ser221.

• The mechanism of catalysis is the same, including the positioning of substrate by hydrogen bonding.

This appears to be a case of convergent evolution: Mother Nature found a good idea a second time.

The Merops database classifies serine proteases (and all others) into clans and families, based on sequence and structure similarity.

Good biochemical sense tells us that the four mammalian proteases diverged, whereas the bacterial one converged.

• The accepted procedure for distinguishing between the convergence and divergence is to count the number of common characteristics:

• If there are many, we are observing divergence, if few, convergence.

In molecular terms, the important similarity criteria for divergence can be listed:

• The DNA sequences of their genes are similar

• Their amino acid sequences are similar

• Their three-dimensional structures are similar

• Their interactions with substrates are similar

• Their catalytic mechanisms are similar

• The catalytic residues are in the same sequence in the chain

These criteria are listed in descending order of importance. If the first two hold, the rest usually follow.

• In the serine proteases, tertiary structure is more conserved than primary structure

[For a review of the evolution of the serine proteases, see: Krem and DiCera, EMBO J., 2001, 12, 3036.]

How is the activation of the zymogens initiated?

• Two peptide hormones are involved, cholecystokinin and secretin.

• The appearance of peptides and fatty acids in the small intestine stimulates the production of cholecystokinin

  This in turn signals for the release of the zymogens, enteropeptidase, NaHCO3, and bile acids from the gall bladder.

• Secretin is released when the pH of the intestine drops below about 5, and it stimulates the further release of pancreatic juices and zymogens.