Every clinical-stage orexin-2 receptor (OX2R) agonist is a small molecule. PeptAI OX2R is developing the first peptid...

Every clinical-stage orexin-2 receptor (OX2R) agonist is a small molecule. PeptAI OX2R is developing the first peptide agonist against this target — and as far as we can determine, the first in active development anywhere.

The receptor and the thesis

OX2R (UniProt O43614) is a Class A GPCR whose natural ligands — Orexin-A and Orexin-B — regulate wakefulness, arousal, and energy balance. Orexin neuron loss is the established mechanism in narcolepsy type 1. A potential role in attention-regulation is an emerging hypothesis (one 2025 clinical study in drug-naive children — this is a thesis, not established biology). The three clinical-stage OX2R agonists (TAK-861, ORX750, BP1.15205) all target narcolepsy and are all small molecules. The peptide modality is open.

The candidates

• OX2R-004 — KGDRYGVAYEHGGAQPFK-NH₂, 18-mer natural-AA agonist designed against the orthosteric binding pocket (PDB 7L1U). Track A reference compound. Principal liability: short predicted plasma half-life (~5 min) — a class limitation of unmodified peptides, not a sequence-specific failure.
• OX2R-004-S1 — Stabilized variant: N-acetylation + D-Arg4 + D-Tyr5 at the primary cleavage sites. Predicted half-life ~55 min (~11× improvement). Sub-systemic threshold routes this candidate to intranasal or subcutaneous delivery rather than systemic.

Last milestone closed — April 17, 2026

Pre-IND Computational Feasibility Assessment v2.1: no-kill verdict. Solubility, aggregation, immunogenicity, and toxicity all clean across both candidates. OX2R-004-S1 advances with two named open challenges: proteolytic stability is conditionally rescued (55 min, still sub-systemic), and OX2R/OX1R receptor selectivity is unresolved. ~100% orthosteric conservation between the two receptor subtypes is a class-level problem — native Orexin-B itself achieves only modest subtype selectivity (~12×). This is a named engineering challenge, approached openly. Three structural gates (binding pose G3 provisional, affinity G4, MD stability G5) remain open pending compute. Full report published on Molecule Labs.

Next milestone

First measured data point: functional EC50 (human Gq Ca-flux assay, orexin-B as reference) on both candidates. Path: scoped assay quote expected this week → purchase order mid-to-late June → data late July / early August. Two honest dependencies: quote landing clean and spend approval following promptly. Either slips, we'll say so.

Worth stating plainly: we have no measured binding or functional data yet. Computational predictions informed the decision to advance — they are not results. The assay is the only test that is actually true.

Cadence

Updates when there is something to report: a gate closes, data lands, or a decision is taken. Next expected update: functional assay results (target late July / early August).