IDE and GLP-1 Degradation: A New Stability Pathway
What happens to GLP-1 after it circulates in the body has largely been told as a story of one enzyme: DPP-4, the dipeptidyl peptidase that truncates the incretin hormone within minutes. Pharmaceutical GLP-1 receptor agonists are engineered to resist DPP-4 cleavage — semaglutide’s amino acid substitution at position 8 is the most widely known example.
A paper published last week in Science Advances adds a second chapter. Researchers identified insulin-degrading enzyme (IDE) as a previously unrecognized protease that cleaves GLP-1 at two distinct sites. The finding opens a new axis for drug design, particularly for analogs intended to reach tissues beyond the circulation.
The researchers began with a targeted proteomics screen, looking for tissue-resident enzymes that could process GLP-1. IDE emerged as a candidate. In vitro assays confirmed that IDE cleaves GLP-1 at two sites — one between residues 14 and 15, another near the C-terminus. Both cleavages inactivate the peptide.
To test whether IDE-mediated degradation matters in a physiological context, the team engineered GLP-1 and semaglutide with D-amino acid substitutions at the two IDE cleavage sites. D-amino acids are mirror-image versions of the natural L-amino acids; proteases that recognize specific L-amino acid sequences generally fail to cut their D-isomer equivalents. The engineered peptides — D-Ser¹⁴ and D-Arg¹⁸ substitutions, alongside the existing DPP-4–protective modifications — showed markedly enhanced stability when exposed to plasma, liver secretomes, intestinal secretomes, and peritoneal fluid.
The more striking finding was central nervous system stability. GLP-1 receptor agonists have historically struggled to maintain stability in CNS tissue, partly because a different set of proteases operates there. The IDE-resistant analogs remained intact in CNS homogenates, suggesting that the same enzyme responsible for peripheral degradation also limits GLP-1 residence time in the brain.
The study’s implications are mechanistic rather than clinical at this stage. No new drug candidate was announced; the D-amino acid variants are proof-of-concept tools. But the identification of IDE as a GLP-1 protease — and the demonstration that dual DPP-4 and IDE resistance can be engineered in a single peptide — provides a rational framework for designing next-generation agonists with extended pharmacokinetics in both peripheral and central compartments.
For context on the current standard of care and reference protocols, the Research Protocols archive maintains indexed GLP-1 compound records, and Research Stacks covers practical protocol references for the class.