Cells Tag Faulty Glycogen for Destruction, Study Reveals

Glycogen, the branched sugar polymer that stores energy in our cells, has long been treated as an inert fuel depot. A new study published in Nature suggests it is anything but neglected. Researchers led by Matthew C. J. Yip and Felix Randow at the MRC Laboratory of Molecular Biology in Cambridge have discovered that cells run a dedicated quality control system over their glycogen granules, and that this system depends on a surprising molecular player: RNF213, an enzyme better known for its role in blood vessel development and disease. When the system fails, poorly branched glycogen accumulates into insoluble lumps called polyglucosan bodies, which damage the brain. The finding establishes that ubiquitin, the small protein tag cells use to mark damaged proteins for disposal, is also deployed against faulty carbohydrates, expanding the reach of molecular quality control far beyond the protein world.

Biologists have spent decades mapping how cells police their proteins. Chaperones refold misfolded chains, the proteasome shreds irreparable ones, and autophagy devours bulky aggregates. DNA repair pathways similarly guard the genome. But glycogen, a polymer of glucose assembled by glycogen synthase and sculpted by branching enzyme, has lacked an obvious surveillance mechanism. That gap matters clinically. In adults, aberrant, poorly branched glycogen builds up as polyglucosan bodies in disorders such as adult polyglucosan body disease, a condition caused by mutations in the branching enzyme gene GBE1 that can produce progressive weakness, numbness and neurogenic bladder. Until now, no one had identified a cellular machine that actively detects and removes these defective polymers before they crystallize into disease.

The Cambridge team began with genetics. Mice engineered to lack the ligase activity of RNF213, an enormous E3 ubiquitin ligase previously implicated in Moyamoya disease, a rare cerebrovascular disorder, accumulated polyglucosan in specific brain regions: the cerebellum, the pons and the hippocampus. That pattern of accumulation pointed to a protective role in astrocytes, the star-shaped support cells that stockpile much of the brain’s glycogen. The inference was striking. If removing RNF213’s enzymatic activity caused polyglucosan to pile up, then RNF213’s normal job might be to recognize aberrant glycogen and mark it for destruction, much as E3 ligases mark damaged proteins.

To test that idea directly, the researchers built a cell-based system in which cells manufacture polyglucosan, the pathological, sparsely branched form of glycogen. In these cells, RNF213 selectively attached ubiquitin to the abnormal polymer while leaving healthy, well-branched glycogen untouched. That selectivity is the crux of the discovery. Quality control is only useful if it distinguishes good from bad, and here the enzyme appeared to do exactly that, reading the structural state of a carbohydrate granule and tagging only the defective version. The tagged polyglucosan was then engulfed by autophagosomes, the cell’s membrane-bound disposal vesicles, in a process the authors describe as autophagy of ubiquitylated aberrant glycogen.

How does a protein enzyme read the architecture of a sugar polymer? Cryo-electron microscopy provided the answer. Wanda Kukulski’s group at the University of Bern joined the effort to solve the structure of RNF213 bound to maltoheptaose, a linear seven-glucose fragment derived from glycogen. The images revealed that RNF213 carries a carbohydrate-binding module, CBM20, that grips linear oligosaccharides. This detail matters because normal glycogen is densely branched, studded with short side chains, whereas polyglucosan is comparatively linear and sparse. The CBM20 domain, the structure implies, acts as a molecular ruler that prefers the exposed linear stretches characteristic of aberrant glycogen.

The structural work also explained a paradox. When the researchers mutated the carbohydrate-binding surface of CBM20, RNF213 did not lose its activity; it gained it, promiscuously ubiquitylating even physiological glycogen. In other words, the binding domain normally restrains the enzyme, preventing it from attacking healthy glycogen and licensing activity only when the polymer presents the right structural cues. Disrupting that restraint converts a precise quality control factor into a blunt instrument. This regulatory logic, in which substrate recognition gates catalytic output, echoes themes from protein quality control but is here applied to a branched carbohydrate, a conceptual first.

Ubiquitylation is only the first step of disposal. The study traced what happens next through epistasis analysis, which places genes in a pathway by testing how mutations combine. RNF213 acts upstream of LUBAC, a linear ubiquitin chain assembly complex, suggesting a hierarchy in which one E3 ligase initiates tagging and a second elaborates the mark. The resulting ubiquitin coat on polyglucosan then recruits autophagy receptors, including SQSTM1, also known as p62, TAX1BP1 and optineurin. These adaptor proteins physically link the tagged cargo to the autophagy machinery, triggering uptake into autophagosomes and eventual degradation. The network implies that glycogen surveillance, like protein surveillance, is a multi-layered system with division of labor among several enzymes.

The discovery also reframes RNF213 itself. The gene has attracted intense clinical interest because loss-of-function variants cause Moyamoya disease, and the protein’s ligase activity has been linked to lipid droplet biology and innate immunity. Adding glycogen quality control to its portfolio suggests that RNF213 is a broad-spectrum surveyor of cytoplasmic polymers, and it raises the question of whether some Moyamoya pathology might involve glycogen mishandling in brain cells, a possibility the study does not directly establish but which now invites investigation. It also broadens the emerging concept of non-protein ubiquitylation, in which ubiquitin marks lipids and carbohydrates, into a central pillar of cellular housekeeping.

For medicine, the implications are tantalizing but early. Adult polyglucosan body disease and related glycogen storage disorders currently lack therapies that address the root problem of insoluble polymer accumulation. If the RNF213 pathway can be understood well enough to enhance, perhaps by stabilizing the enzyme’s activity or boosting the autophagy receptors it engages, cells might in principle be taught to clear polyglucosan before it accumulates. The mouse data showing region-specific polyglucosan buildup in the absence of RNF213 ligase activity provide a model for testing such strategies. Caution is warranted: the work identifies a mechanism, not a treatment, and translating ubiquitin biology into drugs has historically proven difficult.

Scientifically, the study’s deepest significance may be conceptual. Quality control, biologists now realize, is not a protein-only affair. DNA, RNA, lipids and now glycogen all fall under the ubiquitin system’s jurisdiction, each with dedicated sensors and disposal routes. The Cambridge and Bern teams have shown that even a humble energy store is inspected, tagged and recycled with the same molecular rigor applied to a misfolded enzyme. Cells, it turns out, do not merely fill their glycogen granules and forget them. They watch them, and when the branching goes wrong, they call in RNF213.

Subject of Research: Ubiquitin-dependent quality control of glycogen by the E3 ligase RNF213

Article Title: Quality control of glycogen through direct ubiquitylation by RNF213

Article References: Yip, M. C. J., Naydenova, K., Otten, E. G., Heatley, A., Moe, A., Anton, L., de los Reyes-Ramírez, L., Jolin, H. E., Langevin, F., Wiacek, M., Franco, C., Bertolotti, A., Kukulski, W., McKenzie, A. N. J., & Randow, F. (2026). Quality control of glycogen through direct ubiquitylation by RNF213. Nature. https://doi.org/10.1038/s41586-026-11139-6

Image Credits: AI Generated

DOI: 10.1038/s41586-026-11139-6

Keywords: glycogen, RNF213, ubiquitylation, autophagy, polyglucosan, E3 ligase, cryo-EM, CBM20, astrocytes, neurodegeneration, LUBAC, SQSTM1