Abstract
Self-labeling protein tags are widely used in advanced bioimaging where dyes with high-photon budgets outperform their fluorescent protein counterparts. Further increasing the emitted photon numbers of dye-tag systems is actively pursued by both new fluorophore chemistry and protein engineering. By scrutinizing the protein microenvironment of fluorophores, here we propose that proximal thioether groups negatively affect the photostability of the dye-tag system. We attribute the disparity in photostability of rhodamine dyes on HaloTag, SNAP-tag, and TMP-tag3 to the influence of the inherent thioether linkage within the SNAP-tag and TMP-tag3. This photochemical pathway leads us to further devise tags with higher photostability. We first show that rhodamine dyes on TMP-tag3.1, which employs a proximity-induced SuFEx reaction instead of a thiol-acrylamide addition to replace the thioether adduct, achieve photon budgets comparable to those ligands on HaloTag. We further showcase that by mutating the methionine near the fluorophore pocket, HaloTag: M175L generally gives up to four times enhancement on photostability when labeled with red and far-red rhodamines. The enhancement of HaloTag modification is demonstrated with single-molecule fluorescence imaging, live-cell fluorescence imaging, and voltage imaging. During time-lapse imaging, gradual photooxidation of Met leads to a reduced photobleaching rate, mechanistically supporting the thioether pathway hypothesis. Our findings suggest that thioether editing on self-labeling tags is a general strategy to enhance the photostability of fluorophores for advanced time-lapse imaging techniques.