Sugar molecules called glycans, which attach to proteins and lipids to regulate how they fold and function, also bind to RNA, reports a study posted to bioRxiv this week (bioRxiv 2019, DOI: 10.1101/787614). The finding overturns biochemical dogma and hints at new roles for both RNA and glycans—if the results stand up to peer review and replication by other groups.
“If it’s verified I think it would open up a new avenue of research into the structure and function of such modifications of RNA,” says Richard Cummings, a glycobiologist at Harvard University.
The initial report of glycosylated RNA, published by a team led by Stanford University chemical biologist Carolyn Bertozzi, provides more questions than answers, researchers say. “They have data that are really suggestive, but there are so many questions about how this would occur,” says Laura Kiessling, a chemical biologist at the Massachusetts Institute of Technology.
The discovery stemmed from a serendipitous observation during her lab’s study of glycosylated proteins, says Bertozzi. To trace these proteins, scientists in her lab metabolically label cells with precursor sugars tagged with clickable handles that can be modified with imaging probes. Postdoc Ryan Flynn was using this metabolic tagging technique to label glycoproteins but was surprised to find it had labeled a preparation of what he thought was purified RNA.
“I don’t think anyone else in my lab, having seen this labeled species, would have even thought to look beyond proteins and lipids,” Bertozzi says. However, Flynn had done his PhD with RNA biologist Howard Chang, also at Stanford, so he was primed to think about RNA, and he began to look deeper.
To try to identify the labeled species, Flynn tried to degrade it; proteases didn’t work, and neither did DNases. But RNases did—suggesting that the sugar-tagged molecule likely, somehow, contained RNA.
“At first, none of us believed it—we were like, there’s no framework for that,” says Bertozzi. But the team observed these "glycoRNA" in several different contexts. Flynn made his original observation in cultured human cervical cancer (HeLa) cells, so his next step was to look elsewhere. Flynn observed glycoRNA in human embryonic stem cells, mouse leukemia cells, and hamster ovary cells, as well as in living mice.
The researchers found that the glycans attached exclusively to small RNAs—one type of small RNA in particular, called Y RNA. This class of RNA is known for overseeing quality control of RNA-protein complexes.
To learn more about the composition of glycoRNA, the researchers exposed purified RNA from metabolically labeled HeLa cells to a series of enzymes that cleave off different parts of the glycan. The results suggest that the glycans attached to RNA have been built up through a series of enzymatic steps that normally occur when proteins attach to a particular class of sugars called N-glycans.
Bertozzi and Flynn’s experiments so far don’t address questions about how and where in the cell RNA might get tagged with sugars. Proteins mostly acquire their glycan tags in the so-called secretory pathway of the cell—the endoplasmic reticulum and Golgi apparatus, where a subset of newly made proteins are folded, glycosylated, and packaged in vesicles for transport to organelles or the cell membrane for secretion. But RNA is not known to be present in these compartments.
“It’s very surprising to think that RNA—which should not really meet this glycosylation machinery—has glycans on it,” says Hiren Jitendra Joshi, a glycobiologist at the University of Copenhagen.
Bertozzi speculates that these small RNAs somehow enter the lumen of the endoplasmic reticulum and Golgi apparatus, perhaps with the help of a transporter. Although there’s no known transporter that can perform this maneuver, several organisms make transporters that allow RNA to cross membranes, she says.
RNA’s possible trip through the Golgi apparatus and endoplasmic reticulum is “another mystery pathway,” says Harvard’s Cummings. Another possibility to consider, he says, is that RNA doesn’t interact with these organelles at all; glycans may attach to RNA in the cytoplasm.
The study also did not pin down how this attachment happens. Knowing what connects the two biomolecules will help reveal both the mechanism and the function of their interaction, says Joshi. “What would be really nice to see is some direct measurement of the glycan on the RNA, because there are a bunch of different ways this could be explained,” he says. For example, the glycan might associate with a protein that in turn associates with the RNA.
Bertozzi’s lab did a few experiments to try to understand what holds glycoRNAs together. Running the glycoRNA through a sucrose gradient suggested that the molecule connecting the two entities is no bigger than the small RNA itself. At first, the researchers thought that the RNA might be noncovalently associating with the RNA, “but if that’s true it would have to be the most bulletproof, robust noncovalent association I’ve ever seen because we can’t break those two things apart unless we digest with RNase or with an enzyme that cuts off N-glycans from proteins,” Bertozzi says.
“All evidence points to the glycan being somehow connected to the RNA by some covalent network, says Bertozzi. “Whatever it is, it’s not sensitive to proteases.”
The big mystery, of course, is what role glycosylated RNA plays in the cell. So far, the researchers know nothing about its function. A “striking proportion” of the types of RNAs that Bertozzi’s team found to be modified with N-glycans are known to be recognized by the immune system in certain autoimmune diseases, such as lupus, she says. But how such molecules might affect immune regulation is unknown.
The paper caused a stir on Twitter when Bertozzi and Flynn flagged the preprint and requested input from colleagues. The paper will eventually undergo peer review, she says, but “this was so weird that we wanted to just put it out there and get feedback.”
The team now plans probe further into the chemical details of how the glycans and RNAs connect. The first order of business, Bertozzi says, is developing new analytical tools for studying the “weird hybrid biopolymer.”
This story was updated on Oct. 8, 2019, to correct the spelling of Richard Cummings's name.