Using sequence-derived in silico mutagenesis paired with long-run molecular dynamics simulations and binding enthalpy calculations, researchers at University of Limerick designed novel single-molecule peptides that simultaneously engage all three metabolic hormone receptors — GLP-1R, GCGR, and GIPR. The computationally optimized peptides showed more balanced and thermodynamically favorable binding enthalpies across the receptor triad than endogenous hormones or tirzepatide, while proteolytic cleavage sites were strategically substituted to confer metabolic stability.
The obesity-diabetes pharmacology field has progressed from GLP-1R monoagonists (semaglutide) to dual GLP-1R/GIPR agonists (tirzepatide), with triple agonism representing the next logical frontier — a space where retatrutide is already in Phase 3 trials. What distinguishes this work is the fully computational, structure-guided design pipeline rather than iterative medicinal chemistry, potentially compressing early-stage drug discovery timelines. The engineered proteolytic resistance is clinically significant: it could support lower-dose oral formulations, addressing one of the major adherence barriers with injectable GLP-1 therapies. However, critical limitations must be acknowledged: all findings are in silico, with no cell-based receptor activation assays, no animal pharmacokinetics, and no gastrointestinal tolerability data yet. Binding enthalpy predictions, while sophisticated, do not guarantee functional agonism or selectivity in vivo. This is a compelling proof-of-concept for AI-assisted peptide design, but it sits firmly at the hypothesis-generation stage — incremental methodologically, potentially transformative if wet-lab validation follows.
