Background: Understanding the spatial organization of RNA within cells is central to elucidating gene regulation, RNA processing, and localized translation. Subcellular RNA localization patterns underpin fundamental processes such as cell polarization, differentiation, and response to stress. However, capturing these patterns with high specificity and resolution in live cells remains a technical challenge. Among the most promising recent developments in this area is APEX2-mediated proximity labelling,1 which allows researchers to map RNAs or proteins near genetically targeted enzymes with high spatial and temporal resolution. In proximity labelling, the engineered peroxidase APEX2 catalyses the one-electron oxidation of substrates such as biotin-phenol (BP) in the presence of hydrogen peroxide (H₂O₂), generating short-lived radicals that covalently tag nearby molecules. Initially developed for proteomic labelling (Fig A), the technique was later adapted for RNA profiling (e.g. APEX-Seq2), which revealed subcellular transcriptomes in distinct compartments such as the mitochondrial matrix, ER membrane, and nuclear pore. In APEX-Seq, RNA is labelled directly through reaction of BP-derived radicals with electron-rich guanine nucleobases.
However, despite its strengths, current APEX-based RNA labelling suffers from serious limitations. Standard protocols require high concentrations of H₂O₂ (typically 1 mM)3 to activate labelling, which induces significant levels of oxidative stress and causes DNA and RNA damage.2,4 This not only compromises cell viability but also risks artefactual changes in RNA localization, expression, and protein–RNA interactions. For instance, stress response pathways can be rapidly induced, altering the very transcriptome under study. Furthermore, significant H₂O₂-induced DNA damage can activate pathways that sequester RNA processing or repair proteins, confounding proximity-based protein or RNA labelling. While some improvements have been made (e.g. using biotin-aniline, or BAn, as a more selective substrate for RNA over protein labelling), these still rely on damaging oxidant conditions, limiting widespread application and reproducibility.4 Thus, there remains a pressing need for more efficient, less damaging APEX substrates, especially tailored for RNA labelling.
Preliminary results: In recent work in the Larrosa and Schmidt laboratories, we have addressed these challenges by developing a series of novel biotin-phenol analogues with improved permeability and reactivity. Two of these analogues (BP2B and BP3B) show dramatically improved performance for protein labelling compared to standard BP, requiring only 0.01 mM H₂O₂ (100 times less than standard protocols), and requiring only minutes of label incubation rather than hours (Fig B). Importantly, we have shown that in contrast to standard H₂O₂ concentrations used in BP labelling (1 mM), our new analogues perform robust labelling without inducing measurable genotoxicity (Fig C).
These findings raise the exciting possibility that, with proper tuning, these low-toxicity probes could also be optimized for RNA labelling, offering a transformative tool for spatial transcriptomics. This would provide cleaner and faster RNA labelling in live cells, enabling better contrast and resolution, and minimizing perturbation of the system under study due to artefactual damage induced by superfluous reactive oxygen species. Our proposal builds on this foundation to extend our optimized APEX chemistry to RNA, with the aim of creating next-generation tools for subcellular transcriptome mapping.