Proteolytic Stability

Proteolytic stability describes how long a peptide survives contact with the enzymes that exist to take peptides apart. The relevant enzymes differ by compartment. In plasma, dipeptidyl peptidase-4 removes the first two residues from chains presenting alanine or proline in the second position, and neprilysin cuts internally. Aminopeptidases and carboxypeptidases nibble from each end. In the gut, pepsin, trypsin and chymotrypsin cleave with high specificity at defined residues, which is the main reason ordinary peptides are inactive when swallowed.

The numbers explain the design history of the incretin class. Native GLP-1 has a circulating half-life of roughly one to two minutes, which makes the natural hormone useless as a medicine however potent it is at its receptor. Every long-acting analog answers that enzymatic problem: substituting the DPP-4 target residue, adding a fatty acid that blocks access and parks the molecule on albumin, or cyclising the chain so no extended conformation is available to grip. The same toolkit covers D-amino acid substitution, N-terminal acetylation, C-terminal amidation and backbone N-methylation.

The judgement this supports is which half-life number to believe. A compound stable in buffer says nothing; stability is measured in serum, in liver or intestinal homogenate, or in vivo, and those results can diverge by orders of magnitude. Stability is also weighed against activity, since modifications that hide a peptide from an enzyme often hide it from its receptor too.

The recurring error is transferring stability data across species and matrices. Rodent plasma has a different protease complement from human plasma, and an impressive rodent half-life can shorten sharply in people. Grey-market listings quoting a stability figure with no matrix, no species and no method attached are quoting nothing at all.