Defining cell surface and glycosylated RNAs.

Defining cell surface and glycosylated RNAs - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Search Log in Primary site navigation Search Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Trends Genet . Author manuscript; available in PMC: 2026 Jun 4. Published in final edited form as: Trends Genet. 2026 Jun 1;42(10):922–936. doi: 10.1016/j.tig.2026.05.002 Defining cell surface and glycosylated RNAs Eddy Tzintzun-Tapia 1, # , Fardin Aryan 1, # , Ryan A Flynn 1, 2, 3, * Author information Article notes Copyright and License information 1 Stem Cell and Regenerative Biology Program, Division of Hematology/Oncology, Boston Children’s Hospital; Department of Pediatric Oncology, Dana-Farber Cancer Institute; Harvard Medical School, Boston, MA, USA 2 Department of Stem Cell and Regenerative Biology, Harvard University, Cambridge, MA, USA 3 Harvard Stem Cell Institute, Harvard University, Cambridge, MA, USA # Equal contribution Authors Contributions R.A.F., E.T.T., and F.A., drafted the manuscript and figures. R.A.F. supervised and obtained funding for the project. * Correspondence: ryan.flynn@childrens.harvard.edu Issue date 2026 Oct. PMC Copyright notice PMCID: PMC13232687  NIHMSID: NIHMS2174775  PMID: 42225521 The publisher's version of this article is available at Trends Genet Abstract The cell surface is a dynamic interface for biomolecular interactions, classically considered to be governed by proteins, lipids, and glycans. Advances in chemical biology and RNA detection have revealed that RNA molecules also reside at the cell surface. We initially identified a class of these resident RNAs as small noncoding RNAs covalently modified with N -glycans, termed glycoRNAs. Here we review the technologies that have enabled the identification of glycoRNAs and the more general concept of cell surface RNAs (csRNAs). We also highlight emerging evidence for the trafficking of RNA to the extracellular space and the mechanisms that anchor RNA to the surface of living cells. These observations have reshaped our understanding of RNA regulatory mechanisms and encourage further investigation into previously unrecognized functions of RNAs. GlycoRNAs and the Emergence of Cell Surface RNA Biology For decades, RNA was thought to be confined to the nucleus and cytoplasm, functioning in transcription, translation, and post-transcriptional regulation. Because glycosylation machinery resides within the endoplasmic reticulum and Golgi apparatus, where it modifies proteins and lipids destined for secretion or membrane insertion [ 1 ], RNA was not expected to participate in glycan-dependent cell-surface biology. This paradigm shifted with advances in chemical biology that enabled selective labeling of glycans in living cells. Approximately 5 years ago, we published the first report of sialylated N -glycosylated RNAs (~6.5 years since reporting on bioRxiv [ 2 ]) using glycan-selective metabolic labeling approaches. Small noncoding RNAs were unexpectedly found to be covalently modified with N -linked glycans, forming a new class of molecules termed glycoRNAs. The initially published study further suggested that glycoRNAs are presented at the cell surface [ 3 ], and more recent work across multiple research groups eventually established the extracellular topology of these RNAs. The discovery of glycoRNAs has prompted intense interest in understanding how RNAs access the secretory pathway, identifying the specific RNA transcripts at the cell surface, and defining how they are physically retained and organized once there. A wide array of biochemical, genetic, imaging, and sequencing-based strategies has since been developed to define cell surface RNAs (csRNAs), each with distinct detection biases, including methods that enrich for glycosylated RNAs, sialylated RNAs, or topologically extracellular RNAs. Other reviews have covered chemical tools for in situ detection of characterized glycoRNAs [ 4 ], as well as analytical strategies for glycoRNA interrogation such as mass spectrometry [ 5 ]. Here, we focus on the strategies that have uncovered specific RNA transcripts as glycosylated or on the cell surface and the resulting biological discoveries from these technologies. The path of RNA into the secretory pathway Implicit with the eventual surface presentation of glycoRNAs is the need for them to cross a lipid bilayer, having initially been transcribed in the topologically nuclear/cytosolic region of the cell ( Figure 1 ). The association with a cell surface can occur through cis-presentation, possibly through the secretory pathway, or trans-presentation possibly due to RNAs from dead cells sticking to the surface of a neighboring cell. While the tools to definitively establish if prevalence or complete absence of either mechanism are not yet available, key evidence for the cis-presentation model has been published and will be discussed. Figure 1. Evidence RNA localization in the secretory pathway. Open in a new tab (left) A schematic highlighting the major steps of classical secretion in mammalian cells. (right) zoomed regions of the pathway where the SIDT1 and SIDT2 proteins have been reported as localized (top right) which could enable translocation, where upon loss of lysosomal ribonucleases (middle right) accumulation of RNA species in the lysosome including long and short transcripts is observed, and where peroxidase-mediated biotinylation of RNA in the lumen of the ER (bottom right) found small noncoding RNAs present in the endoplasmic reticulum (ER) of multiple cell types. Altogether, providing parallel examples of RNA biology occurring in the lumen of the secretory pathway in mammalian cells. Most secreted proteins enter the luminal space of the secretory pathway in the endoplasmic reticulum (ER), and therefore it stands to reason that RNA may be similar. Direct evidence for such access came from proximity labeling studies that profiled RNAs within the ER lumen. Using an ER-targeted peroxidase-based labeling strategy (MERR HRP-seq), Ren and colleagues identified a specific population of RNAs localized within the ER lumen, distinct from cytosolic or membrane-adjacent transcripts [ 6 ]. These ER-enriched RNAs were strongly biased toward short noncoding species, including U snRNAs, Y RNAs, snoRNAs, vault RNAs, SRP RNA, and 5S rRNA. Similar ER-localized RNA profiles were observed across human cell lines and primary rat neurons, suggesting evolutionary conservation and the generality of this across different cell types. The transcripts identified by MERR HRP-seq strongly overlapped with initially [ 3 ] and recently [ 7 ] discovered glycoRNA transcripts, lending support for a model in which RNAs can enter the ER and encounter the glycosylation machinery prior to cell surface display. Due to the biochemical processing of the HRP-based labeling, pulldown, and direct sequencing, this work did not directly interrogate if the RNAs labeled were glycosylated; however other downstream analytical processing could provide this evidence in future work. While ER lumen access provides a plausible route for RNA glycosylation, additional work has revealed that RNA membrane crossing is both selective and regulated within endolysosomal compartments. Using a “topology-aware” lysosome immunopurification strategy coupled to RNase protection, Ray and colleagues profiled RNAs residing within lysosomal lumens and showed that RNA entry into lysosomes is not stochastic [ 8 ]. Instead, specific RNA classes, including 7SL SRP RNA, Y RNAs, tRNAs, and 5′ TOP mRNAs were selectively enriched, whereas bulk cytosolic RNA was largely excluded. Mechanistically, lysosomal targeting was encoded by RNA-protein interactions, with defined cis-acting features such as the SRP Alu domain and 5′ TOP motifs sufficient to drive lysosomal entry. RNA delivery to lysosomes required autophagy, and rapid lysosomal degradation normally obscured this RNA pool, becoming evident only upon loss of lysosomal RNases. Together, these findings establish lysosomes as selective RNA trafficking hubs and demonstrate that RNA membrane crossing can be mechanistically specified rather than passive ( Figure 1 ). Notably, several of these RNA species, particularly SRP RNA and Y RNAs, overlap with RNAs detected across ER lumen, glycoRNA, and cell surface RNA datasets [ 3 , 6 , 9 ], however the overall activity of the lysosome in this work was catabolic towards the discovered transcripts suggesting the lysosome may be part of the degradation machinery for luminally localized RNAs. The above examples provide empirical evidence, outside of glycoRNAs per se, that RNAs are capable of translocation across the lipid membranes enclosing the secretory pathway. In the context of glycoRNAs, in our initial work, we hypothesized that SIDT1 or SIDT2, the human homologs of the C. elegans RNAi transporter SID1, could facilitate membrane crossing for glycoRNA precursors, but did not have evidence for this. Recently, Zhang and colleagues showed that dual knockout of Sidt1/2 abolished glycoRNA detection and phenocopied extracellular RNase treatment in both in vitro and in vivo assays [ 10 ]. This work also demonstrated that RNase treatment of neutrophils modulates adhesion, transendothelial migration, and recruitment to inflammatory sites in vivo. Functionally, this effect was narrowed to glycoRNAs engaging canonical adhesion pathways through interaction with P-selectin on endothelial cells, linking surface RNA display to established immune trafficking mechanisms. These findings provide a genetic link between the presence of SIDT1/2, glycoRNAs, and functional regulation of the immune system. However, the precise cellular site and molecular mechanism by which SIDT1/2 facilitate glycoRNA trafficking remain unresolved. While exciting and having advanced our thinking about these cellular activities, much is left to be discovered and resolved surrounding the purposeful and perhaps selective trafficking of RNA across membranes. For example, the tissue expression of SIDT1 and SIDT2 are not the same, nor are they homogenous across the body, raising questions around the importance of these proteins for RNA trafficking. Further, Zheng and colleagues showed that human SIDT1 and SIDT2 display intrinsic pH-dependent RNA-binding and RNA-dependent oligomerization, indicating regulated activity rather than passive membrane association [ 11 ]. However, their resolved cryoEM conformations did not reveal a “pore-like” architecture for RNA passage, and these authors did not establish a structural basis of the human proteins, in their system, for RNA translocation. Thus, while this work provides structural support for RNA binding and evolutionary linkage to RNA transport, the molecular mechanism and cellular context of SIDT-mediated RNA membrane crossing remain unresolved. The RNAs that end up on the cell surface Identifying RNAs on the cell surface has become of great interest across a variety of biological contexts. Experimental approaches for defining these RNA transcripts are dictated by the molecular tools used, which shape the resulting biological insights. Here we highlight two related concepts and methods: first, those that are glycan-centric and second, those which report only on the topological feature of being on the cell surface ( Figure 2 ). Figure 2. Approaches for capturing glycoRNAs and cell surface RNA. Open in a new tab Schematic showing various methods for enriching RNAs based on glycosylation status or cell surface topology. A) ManNAz-seq and Clier-Seq leverage metabolic labeling via Ac 4 ManNAz to identify sialoglycoRNAs across different size ranges. B) SPCgRNA and TnORNA are solid-phase chemistry approaches that couple GAO oxidation and hydrazide chemistry to capture glycoRNAs. N -glycoRNAs and O -glycoRNAs are preferentially released using distinct glycosidases on immobilized RNAs. C) StCEL-seq employs a siayltransferase to extend a chemically modified sialic acid for later labeling and enrichment. D) rPAL-seq employs a periodate-mediated oxidation and sialidase release to enrich for sialoglycoRNAs. E) AMOUR reports on cell surface topology using an extracellular, T7-based strategy to amplify cell surface RNAs. Glycobiology methods uncover glycoRNAs, a novel glyconjugate found at the cell surface The field of chemical biology has provided researchers with a number of tools to label biomolecules, allowing for the interrogation of specific targets within the complex cellular environment. Leveraging this concept, our initial discovery of glycoRNAs was enabled by peracetylated N -azidoacetylmannosamine (Ac 4 ManNAz), a glycan metabolic reporter that serves as an analog to the sialic acid precursor sugar N -acetylmannosamine [ 3 ] ( Figure 2A ). In that study, small RNAs were extracted from Ac 4 ManNAz-labeled HeLa and H9 cells and click-conjugated to DBCO-biotin for affinity purification of azide-labeled RNAs (ManNAz-seq). Sequencing of these sialic acid-containing glycoRNAs, or sialoglycoRNAs, revealed an enrichment of over 190 RNA transcripts across both HeLa and H9 cells, many of which were snoRNAs, tRNAs, and snRNAs. This discovery of a new glycosylated biopolymer quickly prompted the development of novel approaches for isolating and sequencing glycosylated RNAs. In the following sections, we will discuss subsequent methods leveraging glycan chemistry or binding proteins and their findings. In 2023, the Yang group added to the chemical toolkit for investigating glycoRNAs by reporting a solid-phase chemistry method for glycoRNA purification and sequencing, which they termed solid-phase capture of oxidized N -linked glycoRNAs (SPCgRNA) [ 12 ] ( Figure 2B ). This approach employs galactose oxidase (GAO), an enzyme from the fungus Dactylium dendroides that catalyzes the oxidation of galactosyl-containing glycoforms. In their workflow, they treated total RNA with GAO to oxidize native glycoRNAs containing galactose or N -acetyl galactosamine residues, generating aldehyde groups for selective capture by hydrazine resin. Given the specificity of GAO, an important consideration here is the sialylation status of the target glycans. An additional sialidase treatment may be included to homogenize the ends of the target glycans for better GAO activity. In their method, the next step was to release RNAs from the resin using PNGase F, an amidase enzyme that cleaves N -glycan chains and glycoRNAs [ 13 ]. Released RNAs were prepared for sequencing using a commercial small RNA library kit that performs sequential 3’ and 5’ adaptor ligation followed by cDNA synthesis and PCR amplification. Analysis of glycoRNAs captured from MIA PaCa-2 (pancreatic ductal adenocarcinoma cell line) and hTERT-HPNE (non-cancer cell line) revealed several classes of small noncoding RNAs as glycoRNAs, most of which were rRNA and tRNA, but also included snRNAs, snoRNAs, miRNAs, tRNAs, and Y RNA. However, glycosylated miRNAs represented the largest class of differentially expressed glycoRNAs between the PDAC and non-cancer cell line, suggesting potential dysregulation of glycosylation in specific miRNA transcripts in PDAC. Notably, even while profiling different cell lines, over half of their non-rRNA and non-tRNA glycoRNA hits were present in the ManNAz-seq glycoRNA dataset. In their work on probing the role of glycoRNAs in immune cell function, the Lu group used two orthogonal approaches for purifying glycoRNAs from murine neutrophils [ 10 ]. In the first method, they used Ac 4 ManNAz metabolic labeling and copper-free click chemistry to purify glycoRNAs from HoxB8-derived murine neutrophils. In the second method, they used wheat germ agglutinin (WGA), a lectin with affinity for GlcNAc and sialic acid sugars, to purify any GlcNAc or sialic acid-containing glycoRNAs from both HOXB8-derived murine neutrophils and mouse primary neutrophils. Comparing the glycoRNA sequencing data generated from both methods, they observed many overlapping RNAs, and found rRNAs, snoRNAs, snRNAs and miRNAs as the most represented RNAs in their data. This work further demonstrated evidence for small noncoding RNAs as the primary substrates for RNA glycosylation and opened the use of lectins as a strategy for native glycoRNA interrogation. While much of the focus on glycoRNAs has been directed towards N -glycosylated RNAs, the Yang group developed a method to profile native O -glycosylated RNA species. Using an identical approach to their solid phase chemoenzymatic method, SPCgRNA, purified RNA was first subjected to GAO treatment for purification via hydrazide resin ( Figure 2B ). However, RNA is released from the resin using the enzyme GalNAcEXO rather PNGaseF [ 14 ]. GalNAcEXO is a glycosidase that acts on ɑ-linked GalNAc moieties and efficiently cleaves the Tn antigen, which is a single O -linked GalNAc on serine or threonine residues, thereby liberating Tn-containing O -glycosylated RNAs (TnORNA). However, it is worth noting that GalNAcEXO broadly targets terminal GalNAc residues, so this method may also capture glycoRNAs with other terminal GalNAc-containing glycans besides the Tn antigen. Using the TnORNA method in MIA PaCa-2 and A549 cells, the authors identified tRNAs and snoRNAs as the most abundant Tn-containing O -glycoRNAs in their analysis, suggesting that both N -glycosylated and O -glycosylated RNAs are predominantly small noncoding RNA species. Applying TnORNA to PDAC and normal adjacent tissue revealed a nearly 6-fold enrichment of O -glycoRNAs in PDAC tissue, indicating a correlation between O -glycoRNA upregulation and PDAC pathology. In a study from the Kiledjian group, the authors performed metabolic labeling in HeLa and HEK293T cells with Ac 4 GalNAz, an azide-functionalized analog of N -acetylgalactosamine biased for O -glycan incorporation, and found labeled RNAs that were localized primarily within exosomes [ 15 ]. Using commercial exosome purification kits, the authors isolated exosomes from Ac 4 GalNAz-labeled HeLa cells, purified the associated RNA, conjugated to DBCO-biotin reagents, and performed affinity purification. To sequence their Ac 4 GalNAz-labeled exosomal RNA, they performed poly(A)-tailing with E. coli poly(A) polymerase prior to Oxford Nanopore sequencing. Their analysis revealed several families of small noncoding RNAs, including snoRNAs, vault RNA, and Y RNAs, in addition to longer RNAs such as rRNAs, long noncoding RNAs (lncRNAs), and RNAs mapping to rRNA pseudogenes. Given their labeling strategy and observation of Ac 4 GalNAz-labeled RNAs as primarily exosomal, enzymatic and mass spectrometry-based analyses will be valuable in identifying the glycan structures on these RNAs. Taking a chemoenzymatic approach, Zeng and colleagues used a truncated mutant of the bacterial sialyltransferase CstII as a tool for labeling native sialoglycoRNAs, a strategy they termed sialyltransferase-mediated chemical enzymatic labeling (StCEL) [ 16 ] ( Figure 2C ). CstIIΔ32I53S, derived from the bacterium Campylobacter jejuni , has ɑ2,8-sialyltransferase activity and tolerates incorporation of azide-containing sialic acids, enabling click chemistry-based labeling [ 17 ]. Applying StCEL to isolate glycoRNAs from HCT116 human colorectal carcinoma cells, they then performed PANDORA-seq, a sequencing method that relies on enzymatic removal of RNA modifications to prevent interference of adaptor ligation or reverse transcription of modified small RNAs during library preparation [ 18 ]. This method uses a combination enzymatic treatment with T4 polynucleotide kinase (T4 PNK) and E. coli ɑ-ketoglutarate-dependent hydroxylase (AlkB) to remove 3’ phosphate modifications and remove RNA methylations such as m 1 A and m 3 C, respectively [ 19 ]. Their sequencing analysis revealed that rRNA-derived and tRNA-derived small RNAs represent the majority of glycoRNAs from HCT116 cells. Extending StCEL to primary human tissue samples, they observed an upregulation of sialoglycoRNAs in over 70% of paired tumor and normal adjacent tissue across rectal, esophageal, and lung cancers. This is consistent with observations of increased N -glycoRNAs and O -glycoRNAs in PDAC tissues, suggesting dysregulation of glycoRNA levels as a potential hallmark of cancer pathology [ 13 , 14 ]. In our most recent efforts, Ge et al. developed a strategy that combines an RNA optimized periodate oxidation and aldehyde labeling workflow with sequencing, termed rPAL-seq [ 7 ] ( Figure 2D ). In this approach, total or fractionated small RNA was polyadenylated using E. coli poly(A) polymerase and terminated with deoxyadenosine triphosphate (dATP), preventing potential side reactivity of periodate with a terminal ribose during the labeling steps. Polyadenylated RNA was then treated with sodium periodate to oxidize vicinal diols to aldehydes in sialic acid residues. The oxidized sialoglycoRNAs were then ligated to an aminooxy-containing biotin reagent for purification, then released from the affinity resin by Vibrio cholerae sialidase. This “capture and release” strategy is critical as it adds an important layer of enzymatic specificity to the released RNAs for analysis. Critically, rPAL-seq requires the use of a heat-inactivated sialidase to control for any RNAs released due to non-enzymatic hydrolysis during the sialidase digestion or any RNase contamination of those enzymatic preparations. The released sialoglycoRNAs were reverse transcribed (RT) using an anchored oligo(dT) primer and a template-switching oligo strategy under conditions that favored RT to transit known and potentially novel RNA modifications, with final sequencing on the Illumina platform. Applying rPAL-seq to a panel of 13 human cell lines, U-rich snoRNAs and snRNAs were identified as the most represented glycoRNAs across all cell types. Although many glycoRNA transcripts were observed to be cell type-specific, clustering of the 13 cell lines by RNA glycosylation levels led to distinct groupings of either hematopoietic or non-hematopeotic origin, whereas lineage clustering based on transcript expression levels alone did not produce similar groups. This dataset was also interrogated to explore mutational signatures: potential glycosylation sites may impair reverse transcription and lead to mismatches during sequence alignment. Analysis of mismatch sites revealed a global enrichment at U bases, including a mismatch event at position 20a of tRNA-Asn-GTT—a conserved acp 3 U site in humans [ 20 ]. While mismatch signatures can arise from multiple RNA modifications, this observation is consistent with our previous identification of acp 3 U as a linker between RNA and glycans. In a recent report by Zhu et al., the authors describe a glycoRNA sequencing method termed click chemistry-based enrichment of glycoRNAs (Clier-seq), which builds on the Ac 4 ManNAz metabolic labeling approach we initially developed [ 21 ] ( Figure 2A ). To explore potentially longer glycoRNA species, the authors performed a similar sucrose gradient separation, as was initially implemented on H9 and HeLa cells [ 3 ], and analyzed the region from 0–2000 nucleotides. Following the same biochemical protocol as ManNAz-seq, they identified tRNAs as a major class of glycoRNA across multiple nasopharyngeal carcinoma cell lines and B lymphoma cell lines, including serine, threonine, valine, and lysine tRNAs, as well as threonine and valine pre-tRNAs. They also identified the vault RNA vtRNA2–1, which is interesting considering the vault RNA vtRNA1–1 was identified as a glycoRNA hit using the AMOUR method by Jiang, et al (described below). When analyzing glycoRNA transcripts aligned to unannotated regions of the genome, the authors identified three highly enriched glycoRNA transcripts–MSTRG.7832, MSTRG.5930, and MSTRG.18836–ranging in size from 248–313 nucleotides, which the authors predicted to be lncRNAs via the RNA Coding-Potential Assessment Tool (CPAT) [ 22 ]. This work highlights the presence of lncRNAs in glycoRNA populations in addition to the well-observed small noncoding RNAs. Glycan-independent methods of identifying RNA transcripts on the cell surface The glycoRNA sequencing strategies described in the preceding section are primarily in vitro (performed on RNA after extraction from cells) and depend on glycan chemistries. Although multiple studies have established the cell surface localization of glycoRNAs, the in vitro nature of these strategies limits topological information. In the following section, we will highlight approaches that utilize topologically-aware, cell surface-centric frameworks to identify cell surface RNAs, a term that is often used interchangeably with glycoRNA. However, because these methods are not glycan-dependent, they broadly capture RNAs on the cell surface, including both glycosylated and non-glycosylated RNAs. For this reason, we will treat cell surface RNA (csRNA) and glycoRNA as overlapping but distinct categories for now. An initial effort to generally profile RNAs on the cell surface came from Huang, et al. in the form of a method called Surface-seq, a biochemical approach for the purification of cell membranes and their associated RNAs [ 9 ]. In this method, cell membranes are extracted and transferred onto poly(lactic-co-glycolic acid) nanoparticles. This was designed to retain the native orientation of the plasma membrane on the membrane-coated nanoparticles (MCNPs). Using this material, 3’ RNA adaptor ligation was performed directly on the nanoparticles to preferentially ligate extracellular-facing RNAs. RNA was then purified from the MCNPs and subjected to standard 5’ adaptor ligation and reverse transcription. The authors of the study focused on lncRNAs during their sequencing analysis and identified a set of 17 lncRNA transcripts as candidate membrane-associated extracellular RNAs (maxRNAs) using the Surface-seq approach. These transcripts include the nuclear RNAs NEAT1, SNHG20, and the highly conserved lncRNA MALAT1. The authors go on to perform a cell surface-optimized variation of fluorescence in situ hybridization to validate the presence of MALAT1 RNA on the cell surface. The authors note that their Surface-seq analysis focused on lncRNAs due to prior reports of lncRNA associated with cell membrane functions [ 23 , 24 ], but analysis of possible small noncoding transcripts was not readily explored. Approaching this question from a different perspective, Li, et al. utilized eosin Y as a tool for generating singlet oxygen [ 25 ], a spatially-constrained reactive oxygen species that is reported to crosslink RNA and protein [ 26 ]. Using a wheat germ agglutinin-biotin conjugate, the authors tethered streptavidin-eosin Y to the cell surface and stimulated singlet oxygen production to perform cell surface crosslinking. They then performed orthogonal organic phase separation (OOPS), a previously reported method for isolation of ribonucleoprotein (RNP) complexes that relies on the physical partitioning of RNPs into the interphase during acidic guanidinium-thiocyanate-phenol-chloroform purification. RNAs were subsequently recovered from the RNPs via proteinase digestion. Sequencing libraries were prepared using the following workflow: 3’ end repair, ligation of pre-adenylated 3’ end adaptors, reverse transcription with SuperScript III, PCR amplification, and recovery of the amplified library with AMPure XP beads. In agreement with our observations [ 27 ], the authors performed a genome-wide CRISPR screen and identified heparan sulfate (HS) as a critical factor for the surface presentation of RNAs. To identify HS-dependent csRNAs, the authors applied their cell surface OOPS (SCOOPS) method to wild type and HS-deficient CHO cell lines. Using a log 2 FC cutoff of >2, the authors identified over 2000 protein-coding transcripts and over 100 lncRNA transcripts as putative heparan sulfate-associated cell surface RNAs (hepRNAs). Additionally, a number of snoRNA transcripts were identified as potential hepRNAs, consistent with our reports of snoRNAs as cell surface glycoRNAs. More recently, using intact cells, Jiang and colleagues developed a T7-based strategy to amplify cell surface RNAs, termed in situ amplification of outer membrane surface RNA (AMOUR) [ 28 ] ( Figure 2E ). Following cell fixation, cells are incubated with a dsDNA probe containing a T7 promoter sequence and a random 9-nucleotide ssDNA overhang. The random overhangs allow for promiscuous hybridization to diverse RNA sequences, rather than targeting specific transcripts. After oligo binding to the intact cells, T7 RNA polymerase is then introduced and initiates transcription at the promoter sequence, and upon encountering DNA-RNA hybrids formed from the random overhangs, the T7 continues transcription along the cell surface RNA molecule and amplifies the transcript. Reverse transcription and library preparation is performed on the amplified RNA product prior to sequencing. Applying the AMOUR strategy to HeLa and HEK293T cells, they found that the relative abundance of small RNAs was higher than protein-coding genes, with tRNAs, snoRNAs, and snRNAs representing the majority of the RNAs identified in their sequencing dataset. Comparing the transcripts identified using AMOUR with those reported by Flynn et al., the authors recovered a substantial overlap of previously identified glycoRNA species, providing further evidence of small noncoding RNAs as genuine constituents of the cell surface. To test for the functional significance of maxRNAs, they used an antisense oligo strategy to block specific maxRNAs on the surface of monocytes and test the effects on monocyte adhesion to human umbilical vein endothelial cells (HUVECs). Using a pool of 20 nucleotide antisense oligos targeting different regions of specific maxRNAs, they incubated the probe sets with live monocytes, fluorescently stained the cells with a calcein dye, and seeded them onto a layer of HUVECs. After quantifying monocyte attachment via a plate reader assay, they found that monocytes incubated with antisense probes against CTSS, FNDC3B, and CMTM6 transcripts exhibited reduced monocyte adhesion. These observations, along with the Lu group’s findings showing a reduction in neutrophil adhesion to endothelial cells following csRNA digestion, demonstrate the functional significance of glycoRNAs and csRNAs in cell-to-cell interactions. Collectively, both glycan-centric and cell-surface-centric approaches converge upon a similar cell surface RNA landscape. Across diverse glycan-dependent approaches, including ManNAz-seq, Clier-seq, SPCgRNA, TnORNA, StCEL, Lectin-based sequencing, and rPAL-seq, small noncoding RNAs are consistently enriched as the predominant glycoRNA species, with a particular emphasis on tRNAs, snoRNAs, and snRNAs. Notably, several of these transcript types were also identified by the topology-based strategies SCOOPS-seq (snoRNAs), and AMOUR-seq (tRNAs, snoRNAs and snRNAs), highlighting RNA populations consistently detected across distinct methodologies. At the same time, topology-based approaches report a broader diversity of RNAs, including lncRNAs and mRNAs, suggesting these methods may sample other populations of RNA on the plasma membrane. These methodological differences are highlighted in Table 1 , which compares the glycoRNA and cell surface RNA sequencing strategies discussed in this review ( Table 1 ). Taken together, these shared and distinct features suggest that, despite differences in labeling chemistries, enrichment strategies, and sequencing workflows, small noncoding RNAs represent a core set of RNA species present at the extracellular face of the plasma membrane. Table 1. Comparison of glycoRNA and cell surface RNA sequencing methods. rPAL-seq [ 7 ] ManNAz-seq [ 3 ] GalNAz-seq [ 15 ] SPCgRNA& TnORNA [ 12 , 14 ] StCEL-seq [ 16 ] Lectin-based sequencing [ 10 ] Surface-seq [ 9 ] SCOOPS-seq [ 25 ] AMOUR-seq [ 28 ] Molecular approach Glycan strategy: periodate oxidation Glycan strategy: Ac4ManNAz metabolic reporter Glycan strategy: Ac4GalNAz metabolic reporter Glycan strategy: GAO oxidation Glycan strategy: sialyltransferase-based labeling (Neu5Az) Glycan strategy: WGA affinity capture Cell surface strategy: RNA 3’ adaptor ligation on purified membranes Cell surface strategy: Eosin Y-mediated surface crosslinking Cell surface strategy: T7-based amplification on intact cells Release strategy Sialidase release from neutravidin beads N/A: on-bead PCR Formamide release from streptavidin beads PNGase F or GalNAcEXO release from hydrazide resin TRIzol release from streptavidin beads TRIzol release from WGA agarose N/A N/A N/A Release specificity control Yes (Heat-inactivated sialidase) No No No No No N/A N/A N/A Reverse transcription and PCR prep strategy Poly(A) extension and template-switching reverse transcriptase 3’ pre-adenylated adaptor ligation prior to RT, followed by cDNA circularization 3’ RT adaptor ligation and template-switching reverse transcriptase 3’ and 5’ adaptors ligated prior to RT 3’ and 5’ adaptors ligated prior to ‘ RT 3’ and 5’ adaptors ligated prior to RT 3’ and 5’ adaptors ligated prior to RT 3’ pre-adenylated adaptor ligated prior to RT 3’ adaptor ligated prior to RT, followed by 5’ adaptor ligation prior to ‘ PCR Reverse transcriptase Maxima H minus superscript III Maxima H minus ProtoScript II Reverse Transcriptase QIAseq miRNA RT MMLV RT ProtoScript II Reverse Transcriptase superscript III MMLV RT Unique molecular identifiers Yes Yes No No Yes No No Yes No Mutation signature or motif analysis Yes No No Yes No No No Yes No Open in a new tab Table comparing the main features of the sequencing strategies described in this review. We divide the strategies by molecular approach: whether they are glycan-centric or cell surface centric. For glycoRNA-specific methods, strategies for releasing from affinity resins are listed. Additionally, we report whether enzyme specificity controls are utilized to control for off-target release during enzymatic digestion of immobilized glycoRNAs. Next, we describe the various adaptor ligation strategies across the different methods, which are an important consideration due to the potential PCR biases incurred during reverse transcription of modified RNAs. The following row indicates the use of unique molecular identifiers (UMIs), a molecular barcoding approach for improving quantitative accuracy in sequencing analysis. Lastly, we have highlighted studies that incorporated sequence motif analysis or mutation signature analysis for the prediction of glycoRNA sites. The anchors of RNA on the cell surface Regardless of the chemical modification landscape of RNAs on the cell surface, much of the work noted above and to date has focused on the cell surface-tethered pool of non-intracellular RNAs (not considering in this review, the vast and interesting field of extracellular RNA, exRNA, more generally defined and reviewed extensively elsewhere [ 29 – 31 ]). For proteins, cell surface tethering comes in many forms: encoded transmembrane domains, protein-protein interactions, and protein-glycan interactions, among others [ 32 – 34 ]. Until recently, little was known about the factors that hold RNA to the cell surface, whether these mechanisms extended to their associated cell surface RNA binding proteins (csRBPs) [ 35 ], and why they might be tethered in this way. Our initial report and profiling of csRBPs demonstrated that at least some of the csRBPs were partially dependent on cell surface RNA for their cell tethering [ 35 ], however we did not systematically define the full scope of this mechanism. In an effort to explore more mechanisms of cell surface RNA, Li and colleagues identified a population of cell surface RNAs, termed hepRNAs, whose stable presentation depends on HS [ 25 ]. Further analyses revealed that hepRNAs are not directly membrane-embedded but instead associate with HS indirectly through cell surface RNA-binding proteins, forming proposed RNA, RBP, HS ternary complexes; loss of HS reduced both csRNA and csRBPs. This is consistent with our initial report of the clustered nature of csRBPs and glycoRNAs into nanoscale domains on the surface of cells [ 35 ], but here expands the number of csRBPs defined adjacent to recombinant TLR7 binding on cells (the main tool used for csRNA detection in this work). Transcript profiling indicated that hepRNAs are enriched for mRNAs, lncRNAs, and snoRNAs, suggesting that this anchoring mode preferentially involves longer RNAs, although biochemical assessment of the posttranscriptional modification landscape of these RNAs was not performed. Finally, the sites of hepRNA accumulation were shown to have certain factors whose association was RNase-dependent, including immune-relevant receptors such as CD155/PVR. Examination of PVR-interacting proteins including CD226, CD96, TIGIT, and KIR2DL5 showed that only the binding of KIR2DL5 to cells was also RNase dependent. Our recent efforts had also been focused on the pathways controlling both general csRNAs (using the anti-RNA antibody 9D5) and glycoRNAs (using recombinant Siglec-11), finding that HS biosynthesis was indeed a central regulator of glycoRNA clustering on multiple cell types [ 27 ]. Our genome-scale genetic screen revealed N - and 6- O -sulfation of heparan sulfate chains to be required for stable retention of the glycoRNA and csRBPs complexes (csRNPs) at the cell surface. A parallel screen for binding of the sialic acid lectin MAA-II did not strongly overlap, demonstrating the selective contribution of HS biosynthesis to the csRNP clusters. The HS-dependency was mirrored with rapid (30-minute) heparinase treatment further demonstrating the critical dependency of csRNP clustering on intact HS chains. Given the selective effects of N - and 6- O -sulfation on the csRNP assembly, we hypothesized that classical cell communication mechanisms that normally leverage HSPGs could be influenced by csRNPs. A classical paradigm in angiogenesis is the use of HSPGs as co-receptors to recruit key growth factors like vascular endothelial growth factor A (VEGF-A) to the cell surface [ 36 , 37 ]. This occurs through a C-terminal HS-binding domain. On cells, a pro-angiogenic form of VEGF-A 165 co-localized with glycoRNAs, and indeed their removal potentiated the activity of VEGF-A 165 on endothelial cells. The interaction between glycoRNAs and VEGF-A 165 was narrowed to the C-terminal domain and mutation of arginine residues in this domain produced a VEGF-A mutant that was less sensitive glycoRNAs, enabling the functional application and study of this pathway both in vitro and in vivo, without the need for RNase treatment. Together, this work establishes heparan sulfate–mediated scaffolding as a key principle governing the anchoring, organization, and function of glycoRNAs at the cell surface. In addition to glycan- and matrix-mediated mechanisms, cell-surface RNA may also be retained through association with protein-anchored RNA-binding proteins. In a 2018 study, Gillissen, et al. identified the canonical RNA-binding protein U5 snRNP200 as an extracellularly exposed antigen on the surface of acute myeloid leukemia (AML) cells [ 38 ], an unexpected finding given its essential functions as a spliceosomal factor in the nucleus [ 39 , 40 ]. Knorr and colleagues followed up on this observation, providing further evidence for cell surface localization of U5 snRNP200 on AML cells but not on healthy hematopoietic cells [ 41 ]. Using a genome-wide KO screen, they found that loss of FCGR2A, a gene that encodes the CD32A Fc receptor, led to reduced cell surface U5 snRNP200. Subsequent validation demonstrated that U5 snRNP200 physically associates with and is dependent on the transmembrane receptor CD32A for cell surface presentation. These findings establish a protein-mediated anchoring mechanism in which a transmembrane receptor tethers an RNA-binding protein at the plasma membrane, providing a plausible route for retention of ribonucleoprotein complexes at the cell surface. Mechanistic Insights and Unresolved Challenges in GlycoRNA Biology Since the discovery of pseudouridine (ψ) in the 1950s, over 170 posttranscriptional modifications (PTMs) have been identified in RNA across living organisms, many of which are abundant on noncoding RNAs such as tRNA and rRNA [ 42 , 43 ]. These modifications enable the full suite of functional activities RNA possesses and subject it to complex regulatory systems, influencing the lifecycle, localization, and structure [ 44 ]. The surprising discovery of glycosylated RNAs presented yet another mode of RNA modification, while also offering a new set of possible localizations and interfaces with biology outside the cell. Consequently, the dependence of glycoRNAs on glycosylation machinery, reactivity with glycosidases, and identification of an RNA-glycan linker acp 3 U have expanded the known molecular substrates of glycosylation systems, bridging the fields of RNA biology and glycobiology. The intersection of these two fields necessitated the development of novel strategies to interrogate glycoRNAs. Recent methods have implemented distinct metabolic, chemical, and chemoenzymatic labeling strategies to capture and sequence glycoRNAs. Nevertheless, the discoveries from these efforts across diverse cell and tissue types have largely converged on small noncoding RNAs [ 3 , 7 , 10 , 12 , 14 – 16 , 21 ]. Across several reports, tRNAs, snoRNAs, and snRNAs have been repeatedly observed, suggesting these families of RNAs as the primary substrates of RNA glycosylation. To date, the majority of evidence for a direct chemical linkage between RNA and glycans has been related to the protein N -glycan linkage. More recently, emerging evidence suggests some glycans on RNA may be connected by an O -linkage [ 14 , 45 ]. Sequencing of O -glycoRNAs via the TnORNA method revealed tRNAs and snoRNAs as major classes of O -glycoRNAs, indicating a commonality with N -glycoRNAs [ 14 ]. Focusing their analysis on miRNAs, the authors compared O -glycoRNAs (TnORNA) and N -glycoRNAs (SPCgRNA) from a PDAC cell line and identified 40 shared glycosylated miRNAs across both methods. This observation raises the question of the mechanisms regulating N - and O -glycosylation of specific RNAs and whether multiple glycan structures can be linked to a single RNA transcript. We have also developed evidence for O -glycans on RNA using genetic knockouts of O -glycan biosynthesis pathway members and performing O -glycomics, however no direct evidence of a covalent bond between O -glycans and RNA has yet been established [ 45 ]. Given the particularities of each method for discovery efforts focused on glycoRNAs, it is important to note methodological considerations that can impact glycoRNA identification. RNase contamination in experimental reagents poses a major risk to RNA integrity and affects the reliability of experimental conclusions. It is therefore necessary to ensure RNase-free conditions when developing new glycoRNA methods. Further, commercial enzymes across different vendors or source organisms can have significant differences in desired activity. In our efforts to establish O -glycoRNA labeling with GAO, we found that different sources of commercial Dactylium dendroides GAO led to either variable O -glycoRNA labeling or non-specific labeling in conditions devoid of RNAs [ 45 ], emphasizing the need to validate observed enzymatic activity to eliminate any potential off-target effects. Moreover, recent gel-based analysis of Ac 4 ManNAz-labeled sialoglycoRNA demonstrated some copurification of non-RNA glycoconjugates [ 46 , 47 ], with protein mass spectrometry analysis suggesting contribution from glycoproteins such as LAMP1 in these samples [ 46 ]. These studies underscore the need for rigorous controls and enzymatic treatments in purification strategies to increase the specificity for capturing glycoRNA for biochemical characterization. Analytically, we also believe there is much to be learned by more deeply investigating the sequencing data being generated. In the RNA modification field, it is widely appreciated that during library preparation, the inefficiencies and errors of RT can be harnessed to define where and what chemical modifications an RNA encodes [ 48 – 50 ]. In the glycoRNA field, as noted above, only a small number of the sequencing efforts are designed to enable robust detection of truncated or mutated cDNAs, and even fewer perform analysis of those cDNA-encoded mutations. In the context of our recently developed rPAL-seq, this type of analysis revealed a clear mutational profile enriched with the glycoRNA transcripts at position 20a of tRNA-Asn-GTT, a conserved acp 3 U site, suggesting this as a possible sequencing-based strategy to validate mass spec-based linker discovery, or to nominate novel sites for modifications in more poorly characterized transcripts. Despite exciting progress in the identification and functional characterization of glycoRNAs, their underlying chemical features are a topic in need of further exploration (see Outstanding Questions ). While our group has identified acp 3 U as a site for covalent linkage between RNAs and N -linked glycans, whether other types of linkages exist remain an open question in the field. Additional questions surrounding the chemical nature of glycoRNAs-such as how modification sites are specified and whether RNA transcripts can host multiple glycan structures-remain. Addressing these questions will require continued development of analytical approaches capable of directly interrogating glycoRNA linkages. Understanding these chemical features will be essential for furthering our understanding of how glycosylation influences RNA function. In parallel, imaging approaches using glycan-based strategies have determined the cell surface localization of glycoRNAs [ 51 – 53 ]. However, surface-presented RNAs may not be limited to glycosylated RNA molecules. Indeed, several groups have worked towards profiling the entire landscape of RNAs at the cell surface, independent of glycosylation status. These efforts have identified lncRNA fragments, small noncoding RNAs, and some mRNAs, among others as cell surface RNAs [ 9 , 25 , 28 ]. Commonalities and differences between glycoRNA and csRNA sequencing datasets call for further examination into the relative abundances of glycosylated and non-glycosylated RNA species on the cell surface. More generally, establishing rigorous and easy-to-implement technologies for defining the extracellular topology of living cells will enable the study of this biology across diverse cellular and physiological contexts. Outstanding Questions. In the context of glycoRNAs and their cell surface nature, while key fundamental questions have been at least answered in part, many remain. The list below is not comprehensive, but rather focuses on key questions that could be answered by leveraging sequencing-based technologies reviewed herein, and include: What features make an RNA amenable to glycosylation or cell surface presentation? Are there common or cell type specific pathways for membrane crossing of RNA? How do we establish the absolute copy number of RNAs on the cell surface? How many glycans can be linked to a single RNA? Are all cell surface RNAs glycosylated? If not, are there distinct functions between glycosylated and non-glycosylated csRNAs? Highlights. Emerging evidence from orthogonal technologies reveals RNAs within compartments of the secretory pathway. Glycan-centric sequencing tools can define RNAs with specific types of N- or O-glycans. Topologic-centric sequencing tools can define RNAs on the extracellular surface of intact membranes. Terminology and nomenclature (Box 1). Clarity in the way things are described is critical for consistency across studies and to ensure straightforward understanding. Included in the > 170 known chemical modifications of RNA is queuosine (Q), a hypermodified form of guanosine biosynthesized exclusively by prokaryotes and salvaged by eukaryotic organisms [ 54 , 55 ]. In the cytosol, Q can be further modified with mannose, and galactose, generating mannosyl-queuosine (ManQ) and galatosyl-queuosine (GalQ) respectively, but they are not thought to be incorporated into the secretory pathway [ 55 , 56 ]; These modifications have been so far excluded from the terminology of “glycoRNAs” as described more generally in this review. The nomenclature to date has therefore preferred single sugar modifications to be assigned as PTMs, while larger complex carbohydrate structures (glycans) on RNA have been termed glycoRNAs. Additionally, as discussed in this review, the vast majority of discovery efforts surrounding glycoRNAs have focused on sialic acid modified glycans, thus leading to a more specific and accurate term of sialoglycoRNAs being reported. Finally, with the discovery surrounding a linker between the glycans and RNA via a nitrogen and the modified RNA base acp 3 U, specification of the linkage chemistry (as also reported in glycoproteins) could also be useful by noting “ N -glycoRNA” for those subsets that are connected by an asparagine-like linkage. Acknowledgments We thank Flynn laboratory members for their constructive feedback. This work was supported by grants from the Rita Allen Foundation (R.A.F.), Cancer Research Institute (CRI14279; R.A.F.), National Institute of General Medical Sciences of the National Institutes of Health under award number R35GM151157 (R.A.F.), National Science Foundation Graduate Research Fellowship Program under award number DGE2140743 (E.T.T.) Footnotes Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Competing Interests R.A.F. is a stockholder of ORNA Therapeutics. R.A.F. is a board of directors member and stockholder of Blue Planet Systems. The other authors declare no competing interests. References 1. Krasnova L and Wong C-H (2016) Understanding the Chemistry and Biology of Glycosylation with Glycan Synthesis. Annu Rev Biochem 85, 599–630 [ DOI ] [ PubMed ] [ Google Scholar ] 2. Flynn RA et al. (2019) Mammalian Y RNAs are modified at discrete guanosine residues with N-glycans bioRxiv, 787614 [ Google Scholar ] 3. Flynn RA et al. (2021) Small RNAs are modified with N-glycans and displayed on the surface of living cells. Cell 184, 3109–3124.e22 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 4. Zhang X et al. (2026) Chemical Biology Tools for Decoding the Cell Surface GlycoRNA. Angew Chem Int Ed Engl 65, e26098. [ DOI ] [ PubMed ] [ Google Scholar ] 5. Yu L et al. (2026) Mass spectrometry-based glycomics towards GlycoRNA. Glycoscience & Therapy 2, 100021 [ Google Scholar ] 6. Ren Z et al. (2023) Enzyme-Mediated Proximity Labeling Identifies Small RNAs in the Endoplasmic Reticulum Lumen. Biochemistry 62, 1844–1848 [ DOI ] [ PubMed ] [ Google Scholar ] 7. Ge R et al. (2025) Catch and release of sialoglycoRNAs enables sequencing-based profiling across cellular and extracellular material. bioRxiv DOI: 10.1101/2025.10.04.680438 [ DOI ] [ Google Scholar ] 8. Ray GJ et al. (2025) Lysosomal RNA profiling reveals targeting of specific types of RNAs for degradation. bioRxiv DOI: 10.1101/2025.09.09.674968 [ DOI ] [ Google Scholar ] 9. Huang N et al. (2020) Natural display of nuclear-encoded RNA on the cell surface and its impact on cell interaction. Genome Biol 21, 225. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 10. Zhang N et al. (2024) Cell surface RNAs control neutrophil recruitment. Cell 187, 846–860.e17 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Zheng L et al. (2024) Cryo-EM structures of human SID-1 transmembrane family proteins and implications for their low-pH-dependent RNA transport activity. Cell Res 34, 80–83 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 12. Li J et al. (2023) Novel Approach to Enriching Glycosylated RNAs: Specific Capture of GlycoRNAs via Solid-Phase Chemistry. Anal Chem 95, 11969–11977 [ DOI ] [ PubMed ] [ Google Scholar ] 13. Xie Y et al. (2024) The modified RNA base acp3U is an attachment site for N-glycans in glycoRNA. Cell 187, 5228–5237.e12 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 14. Li J et al. (2025) O-Glycosylated RNA Identification and Prediction by Solid-Phase Chemoenzymatic TnORNA Method and PONglyRNA Tool. Anal Chem 97, 16906–16915 [ DOI ] [ PubMed ] [ Google Scholar ] 15. Sharma S et al. (2025) Extracellular exosomal RNAs are glyco-modified. Nat Cell Biol 27, 983–991 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 16. Zeng G-D et al. (2025) StCEL-based platform enables detection and sequencing of native sialoglycoRNA bioRxiv, 2025.08.06.668864 [ Google Scholar ] 17. Cheng J et al. (2008) Multifunctionality of Campylobacter jejuni sialyltransferase CstII: characterization of GD3/GT3 oligosaccharide synthase, GD3 oligosaccharide sialidase, and trans-sialidase activities. Glycobiology 18, 686–697 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 18. Shi J et al. (2021) PANDORA-seq expands the repertoire of regulatory small RNAs by overcoming RNA modifications. Nat Cell Biol 23, 424–436 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 19. Trewick SC et al. (2002) Oxidative demethylation by Escherichia coli AlkB directly reverts DNA base damage. Nature 419, 174–178 [ DOI ] [ PubMed ] [ Google Scholar ] 20. Takakura M et al. (2019) Biogenesis and functions of aminocarboxypropyluridine in tRNA. Nat Commun 10, 5542. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 21. Zhu N et al. (2026) Transcriptome-wide identification of glycoRNAs by Clier-seq pipeline. Sci. China Life Sci 69, 72–84 [ DOI ] [ PubMed ] [ Google Scholar ] 22. Wang L et al. (2013) CPAT: Coding-Potential Assessment Tool using an alignment-free logistic regression model. Nucleic Acids Res 41, e74. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 23. Block KF et al. (2011) Association of OLE RNA with bacterial membranes via an RNA-protein interaction. Mol Microbiol 79, 21–34 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 24. Lin A et al. (2017) The LINK-A lncRNA interacts with PtdIns(3,4,5)P3 to hyperactivate AKT and confer resistance to AKT inhibitors. Nat Cell Biol 19, 238–251 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 25. Li Z et al. (2025) Cell-surface RNA forms ternary complex with RNA-binding proteins and heparan sulfate to recruit immune receptors. Mol Cell 85, 4633–4650.e11 [ DOI ] [ PubMed ] [ Google Scholar ] 26. Luo H et al. (2022) Photocatalytic Chemical Crosslinking for Profiling RNA-Protein Interactions in Living Cells. Angew Chem Int Ed Engl 61, e202202008. [ DOI ] [ PubMed ] [ Google Scholar ] 27. Chai P et al. (2026) GlycoRNA complexed with heparan sulfate regulates VEGF-A signalling. Nature 651, 808–818 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Jiang X et al. (2026) Deciphering the RNA landscapes on mammalian cell surfaces. Protein Cell 17, 113–126 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 29. Kolodny GM et al. (1972) Secretion of RNA by normal and transformed cells. Exp Cell Res 73, 65–72 [ DOI ] [ PubMed ] [ Google Scholar ] 30. Valadi H et al. (2007) Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells. Nat Cell Biol 9, 654–659 [ DOI ] [ PubMed ] [ Google Scholar ] 31. Gruner HN and McManus MT (2021) Examining the evidence for extracellular RNA function in mammals. Nat Rev Genet 22, 448–458 [ DOI ] [ PubMed ] [ Google Scholar ] 32. Paulick MG and Bertozzi CR (2008) The glycosylphosphatidylinositol anchor: a complex membrane-anchoring structure for proteins. Biochemistry 47, 6991–7000 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 33. Zhang F et al. (2019) Comparison of the Interactions of Different Growth Factors and Glycosaminoglycans. Molecules 24, 3360. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 34. Hegde RS and Keenan RJ (2022) The mechanisms of integral membrane protein biogenesis. Nat Rev Mol Cell Biol 23, 107–124 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 35. Perr J et al. (2025) RNA-binding proteins and glycoRNAs form domains on the cell surface for cell-penetrating peptide entry. Cell 188, 1878–1895.e25 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 36. Ferrara N et al. (2003) The biology of VEGF and its receptors. Nat Med 9, 669–676 [ DOI ] [ PubMed ] [ Google Scholar ] 37. Sarrazin S et al. (2011) Heparan sulfate proteoglycans. Cold Spring Harb Perspect Biol 3, a004952. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 38. Gillissen MA et al. (2018) AML-specific cytotoxic antibodies in patients with durable graft-versus-leukemia responses. Blood 131, 131–143 [ DOI ] [ PubMed ] [ Google Scholar ] 39. Lauber J et al. (1996) The HeLa 200 kDa U5 snRNP-specific protein and its homologue in Saccharomyces cerevisiae are members of the DEXH-box protein family of putative RNA helicases. EMBO J 15, 4001–4015 [ PMC free article ] [ PubMed ] [ Google Scholar ] 40. Laggerbauer B et al. (1998) The human U5–200kD DEXH-box protein unwinds U4/U6 RNA duplices in vitro. Proc Natl Acad Sci U S A 95, 4188–4192 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 41. Knorr K et al. (2023) Systematic evaluation of AML-associated antigens identifies anti-U5 SNRNP200 therapeutic antibodies for the treatment of acute myeloid leukemia. Nat Cancer 4, 1675–1692 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 42. Cohn WE and Volkin E (1951) Nucleoside-5′-Phosphates from Ribonucleic Acid. Nature 167, 483–484 [ Google Scholar ] 43. Sordyl D et al. (2026) MODOMICS: a database of RNA modifications and related information. 2025 update and 20th anniversary. Nucleic Acids Res 54, D219–D225 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 44. Nachtergaele S and He C (2017) The emerging biology of RNA post-transcriptional modifications. RNA Biol 14, 156–163 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 45. Porat J et al. (2024) O-glycosylation contributes to mammalian glycoRNA biogenesis. bioRxiv DOI: 10.1101/2024.08.28.610074 [ DOI ] [ Google Scholar ] 46. Kegel NB et al. (2025) Proteins are a source of glycans found in preparations of glycoRNA. Exp Mol Med 57, 2505–2516 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 47. Kim S et al. (2025) RNA-associated glycoconjugates highlight potential ambiguities in glycoRNA analysis. Exp Mol Med 57, 2619–2628 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 48. Carlile TM et al. (2014) Pseudouridine profiling reveals regulated mRNA pseudouridylation in yeast and human cells. Nature 515, 143–146 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 49. Werner S et al. (2020) Machine learning of reverse transcription signatures of variegated polymerases allows mapping and discrimination of methylated purines in limited transcriptomes. Nucleic Acids Res 48, 3734–3746 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 50. Motorin Y and Helm M (2024) General Principles and Limitations for Detection of RNA Modifications by Sequencing. Acc Chem Res 57, 275–288 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 51. Liu H et al. (2024) In Situ Visualization of RNA-Specific Sialylation on Living Cell Membranes to Explore N-Glycosylation Sites. J Am Chem Soc 146, 8780–8786 [ DOI ] [ PubMed ] [ Google Scholar ] 52. Ma Y et al. (2024) Spatial imaging of glycoRNA in single cells with ARPLA. Nat Biotechnol 42, 608–616 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 53. Fan S et al. (2025) Molecule Differentiation Encoding Microscopy to Dissect Dense Biomolecules in Cellular Nanoenvironments below Spatial Resolution. Angew Chem Int Ed Engl 64, e202425136. [ DOI ] [ PubMed ] [ Google Scholar ] 54. Harada F and Nishimura S (1972) Possible anticodon sequences of tRNA His, tRNA Asm, and tRNA Asp from Escherichia coli B. Universal presence of nucleoside Q in the first postion of the anticondons of these transfer ribonucleic acids. Biochemistry 11, 301–308 [ DOI ] [ PubMed ] [ Google Scholar ] 55. T S et al. (2025) Biogenesis and roles of tRNA queuosine modification and its glycosylated derivatives in human health and diseases. Cell chemical biology 32 [ DOI ] [ PubMed ] [ Google Scholar ] 56. Kasai H et al. (1976) Letter: The structure of Q* nucleoside isolated from rabbit liver transfer ribonucleic acid. J Am Chem Soc 98, 5044–5046 [ DOI ] [ PubMed ] [ Google Scholar ] ACTIONS View on publisher site PDF (1.1 MB) Cite Collections Permalink RESOURCES Similar articles Cited by other articles Links to NCBI Databases Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top

Related documents

• Engineering polymeric RNA scaffolds as programmable combinatorial innate immune agonists.#13627
• Consensus statement from the second RdRp Summit: towards a unified framework for RNA virus biology#342173
• RNA degradosomes exist in vivo in Escherichia coli as multicomponent complexes associated with the cytoplasmic membrane via the N-terminal region of ribonuclease E#646563
• Editorial: Recent advancements in RNA technologies, diagnostics, and therapeutics#1090553
• Site-Specific RNA-protein covalent bioconjugation using tRNA guanine transglycosylase (TGT)#1157683