The modified RNA base acp3U is an attachment site for N-glycans in glycoRNA

Abstract

GlycoRNA consists of RNAs modified with secretory N-glycans that are presented on the cell surface. Although previous work supported a covalent linkage between RNA and glycans, the direct chemical nature of the RNA-glycan connection was not described. Here, we develop a sensitive and scalable protocol to detect and characterize native glycoRNAs. Leveraging RNA-optimized periodate oxidation and aldehyde ligation (rPAL) and sequential window acquisition of all theoretical mass spectra (SWATH-MS), we identified the modified RNA base 3-(3-amino-3-carboxypropyl)uridine (acp³U) as a site of attachment of N-glycans in glycoRNA. rPAL offers sensitivity and robustness as an approach for characterizing direct glycan-RNA linkages occurring in cells, and its flexibility will enable further exploration of glycoRNA biology.

Keywords: RNA modifications; acp3U; cell surface; glycoRNA.

Figure 1.. Selective oxidation and aldehyde labeling enables sensitive detection of native sialoglycoRNAs.

(A) Cartoon of a sialoglycoRNA highlighting the glycan-RNA linker region which has not been previously defined. (B) Schematic of sialic acid labeled strategies including Ac₄ManNAz labeling (left) or rPAL labeling (right). (C) RNA blotting of HeLa total RNA labeled and detected with Ac₄ManNAz and Dibenzocyclooctyne-PEG4-biotin (DBCO-biotin) or native sialic acids paired with rPAL labeling. In gel detection of total RNA with SybrGold (Sybr, bottom) and on membrane detection of biotin (Streptavidin-IR800, top) is shown. The region labeled “sialoglycoRNA” was quantified. (D) Quantification of data in (C). Each datapoint (biological triplicate) is displayed with the standard error of the mean (S.E.M.) and P values representing unpaired t-tests. (E) RNA blotting of HEK293, 4188, HUH7, and HAP1 total RNA labeled with rPAL as in C. (F) RNA blotting of HeLa total RNA that was treated in vitro with no enzyme, an RNase A and RNase T1 cocktail, or Sialidase labeled with rPAL as in C. Each datapoint (biological triplicate) is displayed with the SEM. (G) RNA blotting of total RNA from HeLa cells treated with various chemical inhibitors. Inhibitors were added to cells in complete media for 24 hours: 0.05% DMSO, 200 μM P-3_FAX-Neu5Ac, 10 μM NGI-1, or 4 μM Kifensuesine, after which RNA was collected and processed for rPAL labeling. Each datapoint (biological triplicate) is displayed with the S.E.M. and statistical analysis was performed using an unpaired t-test. (H) RNA blotting of total RNA labeled with rPAL from four sorted populations of human PBMCs including CD19 (B cells), CD3 (T cells), CD14 (monocytes and macrophages), and CD16 (NK cells, monocytes, and neutrophils) positive cells; RNA from approximately 150,000 cells was used for each lane.

Figure 2.. 3-(3-Amino-3-carboxypropyl)uridine (acp³U) is an endogenous template for RNA glycosylation.

(A) Schematic of the rPAL labeling, enrichment, RNase digestion, on-bead PNGaseF release (top, Release Type A) or on-bead sialidase release followed by in solution PNGaseF digestion (bottom, Release Type B) and SWATH-MS analysis. (B) Heatmap analysis of the nucleosides identified by SWATH-MS from HEK293 and K562 cells after the RNase digestion. Z-scores were calculated for each nucleoside between samples and colored from −1.0 to 1.0. (C) Upset plot intersection of the nucleosides identified by SWATH-MS from HEK293 and K562 cells from both the RNase digestion as well as the on-bead PNGaseF release experiments. (D) Extracted Ion Chromatogram (EIC, specific m/z shown) from the liquid chromatogram (LC) of acp³U-STD, Type B digestion with light (¹⁶O) water, and Type B digestion with heavy (¹⁸O) water. (E) Tandem mass spectrum (MS/MS) analysis of the three peaks shown in (D). m/z values for the two major peaks in each trace are annotated. (F) Chemical structures of relevant molecules. (G) Schematic of the Release Type C. (H) EIC (specific m/z shown) from the LC of acp³U-GlcNAc-STD, Endo F2+F3 released material from rPAL enriched RNA, and a mixture of these two samples (mix). (I) MS/MS analysis of two peaks shown in (H). m/z values for the three major peaks in each trace are annotated.

Figure 3.. The acp³U biosynthetic pathway contributes to glycoRNA production.

(A) Model of a sialoglycoRNA. (B) RNA blotting of U2OS total RNA that was subjected PNGaseF digestion after isolation, rPAL labeling, in gel detection of total RNA (Sybr) and on membrane detection of biotin (Strep). (C) Quantification of the sialoglycoRNA signal in B. Statistical assessment was accomplished by an unpaired t-test, p value shown. (D) Plot profile analysis of the six lanes in B, drawn upwards from the 18S rRNA band in the Strep image using ImageJ. Statistical assessment of the differences in the histogram profiles was assessed using a bootstrapping method (Methods), p values shown. (E) Heatmap plot of z-scores of the nucleoside abundances (normalized for total nucleoside intensity) from small RNA isolated from the three U2OS cell lines and analyzed by SWATH-MS. (F) Quantification of the normalized levels of acp³U in E, t-test performed as in C. (G) rPAL blotting of total RNA samples extracted from U2OS wild type (WT) or two individual knockout clones of the DTWD2 gene. (H) Quantification of the sialoglycoRNA signal from each lane in (G), t-test performed as in C. (I) Plot profile analysis of the nine lanes in G, drawn upwards from the 18S rRNA band in the Strep image using ImageJ. Bootstrapping test performed as in D.

Figure 4.. Proposed biosynthetic pathway for generation of cell surface sialoglycoRNAs.

(A) Cellular tRNAs are synthesized in the nucleus and can be modified there or in the cytosol to create acp³U residue. A subsequent conversion to the carboxamide functionality, by an as-yet unknown enzyme, then allows translocation into the ER lumen. Once in the ER lumen, carboxamide form of acp³U would then enable modification by OST for N-glycosylation. Further trafficking through the secretory pathway accompanied by N-glycan trimming and branch extensions would then produce mature sialoglycoRNAs on the cell surface. Created with BioRender.com.