Decoding Physicochemical Interactions Via Single-Molecule Fluorescence Blinking

Single-molecule fluorescence blinking reflects reversible transitions between open emissive and closed nonemissive forms of rhodamine dyes. These transitions are strongly influenced by the local chemical environment. Here, we establish fluorescence blinking as a quantitative and interpretable readout of local physicochemical interactions. Hydroxymethyl silicon-rhodamine (HMSiR) was covalently linked to a series of short peptides designed to span defined electrostatic, hydrophobic, and hydrogen-bonding properties. Each peptide created a distinct microenvironment that modulated the spirocyclization equilibrium of the fluorophore. Blinking trajectories recorded under controlled conditions yielded descriptors such as on-state dwell times and state-transition statistics, which served as optical signatures of peptide-fluorophore interactions. Machine learning regression mapped these descriptors onto continuous physicochemical parameters, enabling accurate prediction of peptide net-charge, hydrophobicity, and hydrogen-bonding capacity. This work provides a direct connection between blinking dynamics and local physicochemical interactions, transforming stochastic fluorescence blinking into a mechanism-based chemical readout.