Analysis of subcellular transcriptomes by RNA proximity labeling with Halo-seq

Background:

The asymmetrical localisation of RNAs within cells is essential for a range of processes, including cellular development, synaptic function, and signalling (reviewed in (Engel et al. 2020)). This localization is dependent on RNA binding proteins (RBPs), which recognise specific cis sequence elements within RNAs and transport them to the desired subcellular location. Studying RNA distribution in cells has historically relied on direct imaging of small numbers of RNA transcripts using hybridisation-based techniques (Taliaferro 2019), although detection of short RNAs (such as snoRNAs) is often not possible using this method. Recent developments in RNA localisation studies use high-throughput sequencing methods to identify RNA transcripts in neuronal compartments, which can identify hundreds to thousands of transcripts (Cajigas et al. 2012) (Taliaferro et al. 2016). Although this is a major advancement in terms of the number and species of RNA detected, it is only suitable for neuronal cells with long processes that can easily be isolated from the rest of the cell.

The development of RNA proximity-based labelling strategies can overcome several of these limitations. These techniques rely on the generation of reactive oxygen species (ROS) by enzymes, which diffuse away from the protein marker and label RNA species in the vicinity. Marks left by the ROS  allow the purification and identification of the labelled RNA, which can be compared with the bulk total RNA to determine which RNAs are localised. Here, Engel et al. describe a novel proximity-based labelling strategy, Halo-Seq, which uses expression of the HaloTag domain on a protein-of-interest to characterise subcellular transcriptomes with impressive spatial resolution (Engel et al. 2021). 

Findings:

HaloTag is a peptide which forms an irreversible covalent bond with a range of HaloTag ligands (Los et al. 2008). By fusing the HaloTag to a protein which is restricted to a subcellular region of interest, the Halo-seq ligand (dibromofluorescein (DBF)) is also confined to a single area. After irradiation with green light, DBF produces ROS which diffuse away from the point of generation in a 100 nm radius. ROS oxidate RNA bases, which then undergo nucleophilic attack by propargylamine (PA), causing alkynylation. Lastly, using in vitro Cu(I)-catalyzed azide-alkyne cycloaddition “Click Chemistry”, these alkynylated RNA molecules can be selectively purified from the total cellular RNA by the addition of biotin-azide, which reacts with the alkyne group and facilitates the purification of RNA using streptavidin (Figure 1).

Figure 1: A schematic overview of RNA transcriptome characterisation using Halo-Seq. The process can be split up into 3 major steps; labelling of RNA by addition of DBF and irradiation, click-chemistry and separation of labelled RNA, and high-throughput sequencing.

To validate their labelling strategy, the authors cloned a HaloTag domain on histone H2B and NF-kappa B subunit p65 to investigate RNA localisation in the nucleus and cytoplasm respectively. These constructs were then transfected into HeLa cells and the subcellular distribution of each fusion protein was validated through imaging using a fluorescent Halo ligand. RNA dot blots were also carried out to determine whether RNA-Seq can create biotinylated RNA. This assay showed that Halo-p65 and H2B-Halo cells can produce biotinylated RNA in a DBF-dependent manner, indicating that the Halo-Seq technique works within the modified cell lines.

The ability of Halo-Seq to characterise transcriptomes was then investigated. This was done by analysing RNA abundances before and after streptavidin-pulldowns, with localised RNAs being enriched in the pulldown sample compared to the input sample. After sequencing, the differential expression analysis between the input and pulldown samples was carried out using DESeq2. In both the Halo-p65 and H2B-Halo cells, the input and pulldown samples were significantly different, with Halo-p65 samples showing enrichment for cytoplasmic RNAs (such as GAPDH mRNA) and H2B-Halo samples showing enrichment for nuclear RNAs (RNase P, TERC, 7SK RNA). These results validate that Halo-Seq can accurately isolate distinct subcellular populations and characterise nuclear and cytoplasmic transcriptomes. Halo-seq was shown to be able to detect a range of RNA species, including lncRNAs, snRNAs, unspliced RNAs and antisense RNAs.

Additionally, Engel et al. investigated whether Halo-Seq can identify RNA populations in close proximity. HaloTag was fused to the nucleoli marker Fibrillarin, and the previous assay was repeated. When compared with the H2B pulldown, the Fibrillarin pulldown had several nucleoli-specific RNA species enriched (including SNORA68 and 7SL RNA). These RNAs were depleted in the H2B pulldown, which suggests Halo-Seq can distinguish between nucleolar and nuclear RNA populations. Overall, this data indicates that Halo-Seq is a useful tool for transcriptome analysis with impressive spatial resolution.

Subsequently the authors investigated whether any RNA features were present in nuclear or cytoplasmic enriched transcripts. Interestingly, they found that genes with AU-rich elements (AREs) in their 3’ UTR were increased in the H2B pulldown and depleted in p65 pulldowns, indicating that AREs might be associated with nuclear RNA localisation. ARE motifs are bound by RNA binding proteins such as HuR (Lebedeva et al. 2011) and involved with transcript stability and RNA export (Gallouzi and Steitz 2001),. Additionally, the authors note that the number of HuR binding sites in a gene correlates to the degree of enrichment observed in the H2B pulldown.

Following perturbation using leptomycin B (a nuclear export inhibitor), Halo-H2B cells showed enrichment for RNA of 105 genes when compared to leptomycin B negative cells. Further analysis showed that LMB-transcripts have more AREs in their 3’ UTR, with HuR motifs also being enriched. In contrast to a previous study that showed a limited number of RNAs which require HuR for export (Gallouzi and Steitz 2001), this data suggests that the export of hundreds of RNAs are dependent on HuR binding. This work highlights the ability of Halo-seq to quantify dynamic subcellular localisation.

Why I chose this preprint:

The novel Halo-Seq methodology overcomes the major limitations of traditional imaging-based approaches. With better labelling efficiency than other RNA labelling methods that rely on free radical generation (such as Cap-seq (Wang et al. 2019) and APEX-seq (Fazal et al. 2019)), Halo-Seq delivers precise labelling of subcellular compartments with impressive spatial resolution. 

Questions for the authors:

• How did you select DBF as the fluorescent HaloTag ligand for Halo-Seq?
• Will this ligand become available for other labs to carry out Halo-Seq?
• Do you expect Halo-Seq efficiency to vary by cell type?

References:

Engel KL, Arora A, Goering R, et al. (2020) Mechanisms and consequences of subcellular RNA localization across diverse cell types. Traffic 21(6): 404–418. DOI: 10.1111/tra.12730.

Taliaferro JM (2019) Classical and emerging techniques to identify and quantify localized RNAs. Wiley interdisciplinary reviews. RNA 10(5): e1542. DOI: 10.1002/wrna.1542.

Cajigas IJ, Tushev G, Will TJ, et al. (2012) The local transcriptome in the synaptic neuropil revealed by deep sequencing and high-resolution imaging. Neuron 74(3): 453–466. DOI: 10.1016/j.neuron.2012.02.036.

Taliaferro JM, Vidaki M, Oliveira R, et al. (2016) Distal Alternative Last Exons Localize mRNAs to Neural Projections. Molecular cell 61(6): 821–833. DOI: 10.1016/j.molcel.2016.01.020.

Engel KL, Lo H-YG, Goering R, et al. (2021) Analysis of subcellular transcriptomes by RNA proximity labeling with Halo-seq. bioRxiv. DOI: 10.1101/2021.06.08.447604.

Los GV, Encell LP, McDougall MG, et al. (2008) HaloTag: a novel protein labeling technology for cell imaging and protein analysis. ACS chemical biology 3(6): 373–382. DOI: 10.1021/cb800025k.

Gallouzi IE and Steitz JA (2001) Delineation of mRNA export pathways by the use of cell-permeable peptides. Science 294(5548): 1895–1901. DOI: 10.1126/science.1064693.

Lebedeva S, Jens M, Theil K, et al. (2011) Transcriptome-wide analysis of regulatory interactions of the RNA-binding protein HuR. Molecular cell 43(3): 340–352. DOI: 10.1016/j.molcel.2011.06.008.

Wang P, Tang W, Li Z, et al. (2019) Mapping spatial transcriptome with light-activated proximity-dependent RNA labeling. Nature chemical biology 15(11): 1110–1119. DOI: 10.1038/s41589-019-0368-5.

Fazal FM, Han S, Parker KR, et al. (2019) Atlas of Subcellular RNA Localization Revealed by APEX-Seq. Cell 178(2): 473–490.e26. DOI: 10.1016/j.cell.2019.05.027.

Tags: halo, halotag, localisation, proximity, rna, tagging

doi: https://doi.org/10.1242/prelights.30385

Read preprint (No Ratings Yet)