Illuminating the RNA Landscape with Subcellular Profiling

RNA partitioning is critical for cell fate decisions and overall cellular function. Cells sort RNA molecules into specific organelles, such as mitochondria, and membraneless subcellular compartments, including stress granules (SGs).¹ This spatial distribution affects the molecules’ processing, translation, degradation, and storage.² For example, researchers have observed that cells under stress partition particular mRNA species into SGs to regulate their translation and stability.³

Drawbacks of Existing Subcellular RNA Profiling Techniques

Although numerous spatial transcriptomics analyses exist for evaluating RNA localization, these methods face several limitations.

Cell fractionation techniques, such as density gradient centrifugation, separate subcellular compartments based on their different physiochemical properties, enabling subsequent RNA sequencing of the desired fractions.⁴ However, researchers cannot purify all compartments through fractionation, and subcellular compartments of similar density can co-fractionate. These methods also necessitate the use of millions of cells.⁵

High-throughput imaging techniques, including multiplexed error-robust fluorescence in situ hybridization, employ probes to examine the distribution of thousands of RNA molecules simultaneously.⁴ But optical constraints limit the total number of RNA species that scientists can visualize at one time. Moreover, these techniques often require highly trained personnel and specialized equipment.

Proximity labeling approaches, such as APEX sequencing, use enzymes fused to targeting peptides or proteins to label RNA molecules within particular subcellular compartments before enrichment and sequencing.² However, these methods require that the cells express engineered fusion proteins.⁶

Advancing RNA Mapping with a Novel Approach

Haiqi Chen, a reproductive biologist and biotechnology researcher from the University of Texas Southwestern Medical Center, and his team developed a new subcellular transcriptomics-based spatial biology method called photoselection of transcriptome over nanoscale (PHOTON), which combines imaging and sequencing.⁵

PHOTON uses photocleavable primers to bind to RNA molecules within fixed specimens. Following in situ reverse transcription to build the photocaged cDNA library, the researchers labeled the selected subcellular compartments using a fluorescent dye. They then employed a high-resolution microscope and image segmentation pipeline to automatically identify the stained compartments and apply near-ultraviolet laser light to just those regions.⁵ This illumination breaks the photocleavable link attaching the fluorophores to the cDNA molecules, thereby exposing the molecules’ phosphate groups. Moreover, it restores the ability of modified nucleotides in the incorporated primers to form base pairs, known as uncaging. Chen and his group then extracted the nucleic acids and attached PCR handles to the uncaged cDNA molecules only. Following magnetic bead isolation and PCR amplification, they prepared the sequencing library and subjected it to next-generation sequencing.

Uncovering Compartment-Specific RNA Signatures with PHOTON

To identify which RNA species the nucleoli and mitochondria of cultured cells preferentially accumulate, Chen and his team used PHOTON to analyze the molecules within those compartments and in whole cells.⁵ They detected more small nucleolar RNA species in the nucleoli compared to the whole cell transcriptomic data, confirming previous predictions that these RNA molecules localize to the nucleolus. Additionally, the researchers observed the enrichment of several known mitochondrial transcripts within the mitochondrial network. Collectively, these results suggest that PHOTON can accurately assess the subcellular distribution of RNA.

This powerful method also offers scientists an opportunity to examine how subcellular compartments, such as SGs, recruit RNA.⁵ After treating the cells with sodium arsenite to induce SG formation, the team found that long mRNA molecules were more abundant in SGs compared to shorter transcripts, consistent with earlier findings.⁷ Researchers had previously hypothesized that N⁶-methyladenosine (m⁶A) modifications could be important for this length-dependent partitioning. By treating the cells with a drug to decrease the transcripts’ m⁶A methylation, Chen and his group detected fewer long mRNA molecules within SGs. This suggested that the m⁶A marks affect the sorting of RNA into SGs, where this mechanism may play a critical role in how cells respond to stress under normal and disease-related conditions.