GLP-1 receptor blocker (Exendin-4 (3-39))
A lab-made fragment of exendin-4, the lizard-venom peptide behind diabetes drugs like Exenatide, this shortened version blocks the GLP-1 receptor instead of activating it, used only as a lab research tool.
A researcher, an agent, or an algorithm wrote down the sequence and picked a target to hit.
An AI model like OpenFold3 or AlphaFold built a 3D structure and scored how well it fits the binding site.
A second contributor repeated the computation on their own hardware and the scores matched.
A chemistry service or a researcher ordered the sequence, it was manufactured, and mass spectrometry confirmed the right molecule was produced.
A binding or activity measurement confirmed that it actually does what the computer predicted — or didn't.
What this is
Exendin-4 (3-39) is a 37-residue synthetic peptide derived from exendin-4 — the lizard-venom peptide that became the prototype GLP-1 receptor agonist drug exenatide. The fragment removes the first two amino acids (His-Gly) from the N-terminus of full-length exendin-4, leaving the rest of the 39-residue sequence intact. Because the very N-terminus of exendin-4 is the part that triggers receptor activation, truncations in that region are a standard tool in GLP-1 receptor pharmacology research, where they help separate receptor binding from receptor signalling.
The stored sequence is EGTFTSDLSKQMEEEAVRLFIEWLKNGGPSSGAPPPS (37 aa). Compared with full-length exendin-4 (HGEGTFTSDLSKQMEEEAVRLFIEWLKNGGPSSGAPPPS, 39 aa, the parent peptide of exenatide), this fragment is missing only the N-terminal His-Gly dipeptide; the C-terminal nine-residue extension and the rest of the agonist backbone are preserved. The native exendin-4 peptide carries a C-terminal amide (-NH₂) that is not represented in the stored one-letter sequence.
This peptide is a research-tool molecule rather than a clinical drug. It is not approved for any human indication and has no marketing history of its own.
History
Exendin-4 — the parent peptide — was isolated in the early 1990s from the venom of the Gila monster lizard (Heloderma suspectum). The peptide turned out to be a long-lived mimic of human GLP-1, the gut hormone that stimulates insulin release after meals. That discovery led directly to exenatide, the first GLP-1 receptor agonist approved by the FDA (Byetta, 2005), and ultimately to today's GLP-1 drug class (Parkes and colleagues, 2013).
In parallel with drug development, researchers built a toolkit of exendin-4 truncations and modifications to probe how the peptide engages the GLP-1 receptor. The receptor is a class B G protein-coupled receptor with a large extracellular domain that captures the C-terminal half of the ligand and a transmembrane bundle that reads the N-terminus to trigger activation (Donnelly, 2012; Graaf and colleagues, 2016). Removing residues from the N-terminus weakens or abolishes activation, while leaving the C-terminal residues intact preserves the receptor-binding "anchor." Exendin-4 (3-39) sits in this family of N-terminally trimmed analogs used to dissect that two-domain binding model.
What it does
The GLP-1 receptor (GLP-1R) is the target of every drug in the GLP-1 class — semaglutide, liraglutide, tirzepatide, exenatide. Activation of GLP-1R on pancreatic beta cells increases insulin secretion in response to glucose, suppresses glucagon release, slows gastric emptying, and reduces appetite through actions in the central nervous system (Donnelly, 2012).
Exendin-4 (3-39) targets the same receptor but is the N-terminally trimmed form rather than the full agonist. In the two-domain model of class B GPCR binding, the C-terminal portion of exendin-4 docks onto the receptor's extracellular domain to provide high-affinity anchoring, while the N-terminus inserts into the transmembrane core to trigger the activating conformational change (Graaf and colleagues, 2016). Removing the N-terminal residues — as in this (3-39) fragment — is the standard strategy for separating those two steps in research settings. Specific receptor-binding and functional data for Exendin-4 (3-39) itself are not characterised in the references currently attached to this card; what is established in the dossier is the general framework that defines what an N-terminally truncated exendin-4 is for, and how it differs from the active drug.
A separate research interest in the broader exendin-4 family is brain delivery: Salameh and colleagues (2020) reviewed how incretin receptor agonists, including exendin-4 and exenatide, cross the blood-brain barrier and are being investigated as candidate therapeutics for Alzheimer's and Parkinson's disease. That line of work concerns the full-length agonists, not this truncated fragment.
Evidence
- Human: No human clinical trials of Exendin-4 (3-39) are documented in the references attached to this card. The fragment is a research-use peptide, not a clinical investigational drug.
- Animal: No fragment-specific in vivo studies are documented in the dossier references. Parent-peptide (exendin-4 / exenatide) animal pharmacology is extensive and underpins the clinical drug class.
- In vitro: The framework for evaluating exendin-4-derived peptides at GLP-1R via class B GPCR pharmacology is reviewed by Donnelly (2012) and Graaf and colleagues (2016). Fragment-specific binding affinities, EC₅₀ values, or signalling-bias measurements for Exendin-4 (3-39) are not reported in the dossier references.
Mechanism
GLP-1R is a class B (secretin family) G protein-coupled receptor that signals primarily through Gαs, raising intracellular cAMP and activating downstream PKA and Epac pathways in pancreatic beta cells and other GLP-1R-expressing tissues (Donnelly, 2012). The receptor has a large N-terminal extracellular domain (ECD) joined to a seven-transmembrane core. Peptide ligands engage this receptor in two phases: the C-terminal half of the peptide is captured by the ECD with high affinity, and the N-terminal residues are then presented to the transmembrane bundle to trigger the activating conformational change (Graaf and colleagues, 2016).
Exendin-4's exceptional properties as a drug template — high GLP-1R potency and resistance to degradation by dipeptidyl peptidase-4 — derive in part from the C-terminal nine-residue extension (the "Trp cage," residues 31–39 in full-length exendin-4) that anchors strongly to the ECD, and from the substitution of glycine for alanine at position 2 that protects it from DPP-4 cleavage (Donnelly, 2012; Parkes and colleagues, 2013). Exendin-4 (3-39) preserves the C-terminal anchor but removes the first two residues (His-Gly), so the part of the molecule that engages the transmembrane core and drives signalling is altered relative to the full agonist. The detailed pharmacological consequence of this specific (3-39) truncation — partial agonism, antagonism, biased signalling, or another profile — is not characterised in the dossier references and would need to be drawn from primary studies not currently attached to this card.
Related peptides
- Exenatide — the FDA-approved drug. Exenatide is the synthetic full-length exendin-4 (39 aa, sequence HGEGTFTSDLSKQMEEEAVRLFIEWLKNGGPSSGAPPPS), approved as Byetta (2005, twice-daily) and Bydureon (2012, once-weekly microsphere depot) for type 2 diabetes (Parkes and colleagues, 2013).
- GLP-1 (7-37) and GLP-1 (7-36)NH₂ — the endogenous human incretin hormones whose receptor exendin-4 mimics. GLP-1 has only 53% sequence identity with exendin-4 but engages the same receptor by the same two-domain mechanism (Donnelly, 2012).
- Liraglutide, semaglutide, tirzepatide — clinical GLP-1 receptor agonists derived from the human GLP-1 backbone rather than from exendin-4. Liraglutide and semaglutide use fatty-acid acylation for albumin binding and extended half-life; tirzepatide is a dual GIP/GLP-1 receptor agonist (Graaf and colleagues, 2016).
▸full evidence table2 metrics
| metric | value | tool |
|---|---|---|
| ipTM | 0.869692325592041 | boltz-2 |
| ranking score | 0.774983823299408 | boltz-2 |
▸structural qualityopenfold3
| metric | value | note |
|---|---|---|
| gpde | 0.846 | global PDE — lower = better |
| disorder | NaN | fraction disordered |
▸3-letter notation
▸recipeboltz-2 1.0
| parameter | value |
|---|---|
| model | boltz-2 1.0 |
| weights | — |
| hardware | nvidia_nim_api |
| mlx version | — |
| python | — |
| random seed | — |
| msa strategy | none |
| diffusion samples | 1 |
| runtime | — |
| predicted by | mlx@peptide |
| predicted at | 2026-04-24 |
▸citationbibtex
@peptide{pep10532,
sequence = {EGTFTSDLSKQMEEEAVRLFIEWLKNGGPSSGAPPPS},
target = {glp-1r},
author = {peptidemodel},
year = {2026},
status = {synthesized}
}