The mission

Many essential biological and cellular processes require RNA to be in the right subcellular compartment at the right time. Mislocalization of RNA can contribute to the pathologic basis of disease. Tools to assess context-dependent RNA subcellular localization within cells provide a gateway to understanding the relationship between RNA localization and healthy and disease states. Over the past decade, inspired by protein profiling methods1, radical-based RNA proximity labelling tools have been developed to probe the RNA composition in different cellular compartments2,3,4. However, these radical-based approaches tend to label RNAs inefficiently and are likely to induce cellular stress, limiting the use of these methods in the study of dynamic RNA trafficking during stress and physiological processes. A platform that uses non-radical reactive warheads to interrogate RNA subcellular localization is a much-needed tool in the repository of RNA proximity labelling methods.

The solution

Proximity labelling requires the controlled in situ generation of a reactive species when triggered. Because RNA is known to react with carbonyl-based acylating agents5, we sought to mask reactive carbonyl compounds to develop a non-radical reactive moiety for proximity-dependent RNA labelling. To achieve this, we masked the enol tautomeric form of an activated carbonyl compound with a bulky ester that is bioorthogonal and stable in cells, but readily cleaved by the bacterial enzyme BS2 esterase. When BS2-expressing cells are treated with ester-masked probes, the esterase rapidly unmasks the ester, releasing the enol, which quickly tautomerizes to its active keto form, generating the reactive intermediate (Fig. 1a). This intermediate reacts with neighbouring biomolecules or gets quenched, restricting labelling events to a tight radius around BS2. By localizing BS2 to different subcellular compartments, biomolecules in the compartment of interest can be selectively labelled. We synthesized and tested several reactive intermediates and found that masked acid chlorides are optimal RNA proximity labelling agents. A masked acid chloride with an alkyne functional handle, AC-2, emerged as a leading probe.

Fig. 1: Masked acylating agents for proximity labelling of biomolecules.
figure 1

a, A masked acylation agent is unreactive towards biomolecules, as it is trapped in an enol form by a methylcyclopropyl ester, which is biorthogonal to mammalian cells. Localized BS2 esterase unmasks the probe to release an enol, which rapidly tautomerizes to a carbonyl-based acylating agent, such as a thioester or acid chloride. Biomolecules in the vicinity of BS2, such as RNAs, react with the acylating agent and form a covalent bond. The probe includes a click handle to install a fluorescent probe for visualization or an affinity-based enrichment handle for isolation of the labelled molecules. b, Confocal fluorescent images in HEK293T cells with BS2 esterase expressed in various cellular compartments (stained by the fluorescent dye Alexa-555), followed by treatment with AC-2 for 10 min, and then cell fixation and click chemistry to install a fluorescent dye on AC-2 (Alexa-488). There is a strong correlation between BS2 location and labelling. ERM, endoplasmic reticulum membrane; NBD, nitrobenzoxadiazole. © 2024, Pani, S. et al.

Full size image

We integrated AC-2 with BS2 to generate an acid chloride that covalently installs a click-chemistry handle on nearby RNAs, allowing for imaging and next-generation sequencing workflows to create a non-radical RNA proximity labelling platform — bioorthogonal masked acylating agent for proximity labelling and sequencing (BAP-seq). Using confocal microscopy, we validated that labelling is highly localized to BS2, even in nanometre-scale ‘membrane-less’ compartments such as nuclear pores and nucleoli (Fig. 1b), owing to the short half-life of acid chlorides in physiological conditions. Using BAP-seq, we identified many of the known RNAs localized in well-studied compartments, such as the nucleus and mitochondria. In addition, various known and unknown RNAs were found to be localized in the nucleolus, a compartment that is challenging to study because it is not easily isolated. Our BAP-seq datasets validate known RNA constituents of these compartments and provide localization information on numerous previously unannotated RNA species.

Future directions

Although we used the masked acid chloride probes with BS2 to study RNA subcellular organization, the same chemistry can ostensibly be applied to interrogate any biomolecule capable of reacting with electrophiles. The probe’s synthetic design can be adapted to mask different classes of reactive acylating agents. This flexibility offers a route for tuning the reactivity of the intermediate and thereby modulating the labelling radius and optimizing the labelling efficiency for different classes of biomolecule.

BAP-seq is an unbiased platform for profiling the cellular organization of RNA in cells. However, like other proximity labelling methods, labelling is limited to solvent-exposed sites, which could generate false negatives for inaccessible transcripts. Conversely, a highly solvent-accessible transcript would probably be labelled more efficiently than less-accessible counterparts. To tackle some of these issues, BAP-seq data are best interpreted in a relative context, by comparing enrichments from a compartment of interest to those from a control compartment.

BAP-seq is likely to be less perturbing than existing methods of proximity labelling because it does not require the use of hydrogen peroxide or light. Thus, BAP-seq could enable applications beyond well-established steady-state systems, moving towards probing dynamic cellular processes and in vivo applications. With more refined tools, we plan to leverage BAP-seq to generate global, high-resolution maps of dynamic RNA movements during physiological and pathological processes.

Shubhashree Pani & Bryan C. Dickinson

University of Chicago, Chicago, IL, USA.

Expert opinion

“This is an impressive and exciting paper. The bioorthogonal electrophile-activation strategy introduces a new class of diffusible reaction intermediates for microenvironment probing. A similar design philosophy could be applied to develop additional proximity labelling systems. This particular implementation appears to robustly and directly proximity label RNA and, thus, appears to overcome a long-standing need in the field. With further validation, BAP-seq could be a powerful tool for RNA spatial biology.” David Shechner, University of Washington, Seattle, WA, USA.

Behind the paper

In the Dickinson lab, we adopt an interdisciplinary approach to designing molecules with functions to interrogate or modulate biological systems. In our quest to mask the enol of a carbonyl with an ester, we started by designing aryl thioesters to test the concept but soon realized that the reported synthesis routes didn’t yield the desired regioselectivity. After screening reaction conditions, we succeeded in synthesizing the masked acylating probes, but they only labelled biomolecules in live cells to a moderate degree. Each new probe design taught us a lesson on how to make the next one better. Amidst the design–test–design cycle, a breakthrough occurred when we first saw acid chloride-probe labelling in cells by immunofluorescence. The high labelling efficiency and near-perfect colocalization with BS2 indicated the potential of this method. ‘Seeing is believing’ felt quite literal, and, indeed, those imaging experiments presaged the success of the masked acid chloride-based probes. S.P.

From the editor

“Proximity labelling approaches for RNA have relied on chemistries requiring high reagent concentrations and long labelling times owing to their poor reactivity towards RNA. It’s impressive that Dickinson and colleagues developed a robust method for mapping RNAs across different subcellular locations of live cells without these limitations. By leveraging an exogenous subcellularly targeted esterase to unmask highly reactive acylating probes, the 2ʹ-OH group of proximal RNAs can react and be subsequently mapped.” Stacey Paiva, Senior Editor, Nature Chemistry.