Molecular Engineering of Insulin Icodec, the First Acylated Insulin Analog for Once-Weekly Administration in Humans

This is the Novo Nordisk research-group paper that describes the molecular engineering rationale for insulin icodec (Awiqli) — the first once-weekly basal insulin analog approved for human therapeutic use. The molecule was designed against three simultaneous constraints: a plasma half-life long enough to support once-weekly subcutaneous administration, sufficiently reduced insulin-receptor affinity to flatten the glucose-disposal time course across the seven-day dosing interval, and retained subcutaneous-absorption pharmacology compatible with prefilled-pen delivery. The Kjeldsen authors describe the iterative substitution-and-acylation program that produced icodec from a panel of candidate molecules.

The final molecule carries three amino acid substitutions in the native human insulin sequence — Tyr to Glu at A14 (A14E), Asn to His at B16 (B16H), and Phe to His at B25 (B25H) — and one C20 fatty-diacid acylation attached to Lys at B29 (LysB29) via a γ-Glu-OEG-OEG linker (the same linker chemistry that anchors semaglutide to albumin). The fatty-diacid acylation drives strong reversible albumin binding with affinity sufficient to confer a several-day plasma residence and protection from proteolytic and renal clearance; the A14E and B25H substitutions reduce insulin-receptor binding affinity to approximately 0.1% of human insulin, which compensates for the high plasma concentrations the long residence produces and flattens the receptor-occupancy-time profile across the dosing interval; the B16H substitution stabilizes the hexameric self-association state, slowing dissociation to bioavailable monomeric insulin and supporting the multi-day depot kinetics.

The combined molecule has a terminal plasma half-life of approximately 196 hours in human subjects (versus 4–6 minutes for native intravenous insulin and ~25 hours for the previous-generation once-daily degludec). The authors report receptor pharmacology consistent with a true insulin agonist at the IR-A and IR-B isoforms, retained but markedly reduced binding to IGF1R (consistent with the broader safety framework for engineered insulin analogs developed in the analog era), and the absence of mitogenic-bias signals against the conventional reference. The paper anchors the subsequent ONWARDS Phase 3 program (ONWARDS-1 through ONWARDS-6) that produced the regulatory evidence supporting EU approval in 2024 and FDA approval as Awiqli in March 2026 for adult type 2 diabetes after a 2024 complete response letter citing hypoglycemia concerns in type 1 disease.