Together, these data reveal one of potentially many answers to the question “why are RNAs glycosylated?” to which we found that it is to prevent autoinduction of RNA-sensing pathways.

Lastly, we found both TLR3 and TLR7 are required for sensing acp3U containing RNAs.

We teamed up with the Carell lab who synthesized an 18-nt long native or acp3U-containing RNA and found acp3U-RNAs were more immunostimulatory than the unmodified RNAs.

To our surprise, cells deficient in acp3U were not able to stimulate macrophages in both our transfection and efferocytosis assays.

Recently, it was found that acp3U (a hypermodified uridine) is an attachment site for N-glycans to RNA. Genetic KOs of the enzyme DTWD2 prevents the synthesis of acp3U resulting in decreased glycoRNA signal.

Injecting WT mice with deglycosylated apoptotic cells was also capable of inducing a localized RNA-driven innate immune response.

We found that treatment of apoptotic cells with PNGase F elicited a potent IFN response when fed to macrophages.

Next, we checked to make sure PNGase F was not acting on the cells to induce an IFN response. We treated long RNA fractions (containing no glycoRNAs) with or without PNGase F and observed no production of IFNb, confirming this requires deglycosylation of small RNAs.

Here, we harvested small RNA fractions (enriched for glycoRNA) and treated small RNAs with the N-glycosidase PNGase F. We found that removal of the N-glycan renders these RNAs immunogenic upon transfection. Addition of RNases was able to abrogate this response.

In 2021, Ryan Flynn and @carolynbertozzi.bskyverified.social discovered the existence of glycoRNAs which are localized to the cell surface. Since endocytosis is a constitutive cellular function, it is common to find cell surface molecules within the endosomal compartment.