Rational design of a genetically encoded fluorescent biosensor for real-time and bidirectional detection of μ-opioid...

The μ-opioid receptor (MOR) is a key therapeutic target for pain management. However, robust tools for real-time bidirectional monitoring of MOR activation remain limited. Here, we report GEgMOR, a genetically encoded fluorescent biosensor developed via a rational design strategy combining artificial intelligence (AlphaFold3) with genetic engineering. This strategy drastically reduced the number of required mutant constructs from thousands to merely 239, markedly enhancing the efficiency and precision of biosensor development. GEgMOR exhibits outstanding performance in live-cell assays, with a maximum fluorescence response (ΔF/F₀ = 472.8%) 8.6-fold higher than the current state-of-the-art MOR sensor, nanomolar affinity (EC₅₀ = 39.67 nM), rapid kinetics (τ_on ≈ 1.7 s, τ_off ≈ 4.9 s) and high opioid specificity, and excellent photostability under both one-photon and two-photon excitation. Further validation in SH-SY5Y and PC12 neuronal cell models confirmed that the core functional properties of GEgMOR are well maintained in a neuronal cellular context. Notably, GEgMOR enables the first real-time, continuous, and bidirectional monitoring of MOR conformational changes upon agonist activation and antagonist inhibition in a single system. This overcomes the limitations of discontinuous, indirect readouts in conventional stepwise assays. Beyond serving as a powerful dynamic MOR imaging tool, this biosensor establishes a scalable cell-based platform for MOR ligand profiling and drug candidate screening. Our work not only provides a powerful imaging and functional analysis tool for MOR research but also establishes a generalizable strategy to accelerate G protein-coupled receptor (GPCR) biosensor development and advance drug discovery.