Single-molecule Förster resonance energy transfer (smFRET) has established itself as a popular tool in biophysics, offering direct measurements of structural heterogeneity and dynamics. However, smFRET is usually considered a “low-resolution” method in a structural biology context. Although the resolving power, e.g., manifested in the narrowness of an smFRET histogram peak, is theoretically only limited by photon shot noise and can be increased with more photon detections, smFRET experiments on surface-tethered molecules are often subject to extra broadening that drastically limits resolution. Here, we investigate multiple experimental factors influencing smFRET resolution and seek its ultimate achievable limits. We establish direct benchmarks using DNA rulers across different experimental modalities, including conventional surface-immobilized approaches with confocal and prism-TIRF detection, and the recently introduced tether-free ABEL-FRET platform in an anti-Brownian electrokinetic (ABEL) trap. We find that resolution is limited by artifactual heterogeneity in surface-immobilized measurements to various degrees depending on immobilization strategy and detection method. By removing the requirement of surface-immobilization, ABEL-FRET can recover the intrinsic photon-limited resolution while delivering large photon budgets, leading to “ultra-resolution” smFRET. We further find that many commonly used dye pairs produce structured or heterogeneous FRET peaks, precluding ultra-resolution performance. We demonstrate these new ultra-resolution capabilities by resolving subtle distortions to B-form duplex structure induced by common DNA lesions at the single-molecule level. Overall, these results suggest new classes of higher structural-resolution biophysical questions are now available to smFRET.
Whaley-Mayda et al. (Sun,) studied this question.