Molecular glue degraders are more versatile than previously thought

More proteins may be druggable with molecular glue degraders than initially thought, according to a new paper in Science from the biotech firm Monte Rosa Therapeutics. 

Molecular glues are small molecules that act as a connector between a target protein of interest and an E3 ligase, which directs the target to the cell’s protein disposal machinery. The technology has been enjoying a surge of interest as big drug companies start to take notice. But the glues were thought to act only on specific proteins in the cell. 

The study from Monte Rosa scientists challenges that notion. While degraders that recruit cereblon as an E3 ligase were thought to only be able to target proteins with a certain loop motif, Monte Rosa says they can actually target a much wider variety (Science 2025, DOI: 10.1126/science.adt6736).

John Castle, Monte Rosa’s chief data and information officer, says the dogma was that these glues had a limited range of targets. “I think that we’ve blown away those dogmas,” he says, “and in that regard, it’s a watershed moment for the field that you can, with chemical biology, go after these undruggable targets in a way that wasn’t thought to be possible before.”

Essentially, cereblon and its target proteins can change shape to form a complex with the cereblon, the target, and the molecular glue, which enables more options for what complexes are possible and what proteins can be targeted. This includes complexes that don’t use the loop motif.

“It’s just not what people had thought, quite narrowly, about cereblon’s ability to be used through molecular glue degraders,” says Sharon Townson, Monte Rosa’s chief science officer.

One class of degrader that doesn’t require the loop motif is Monte Rosa’s VAV1 degraders, which, by targeting immune-related signaling proteins, are possible therapeutics for autoimmune and chronic inflammatory disorders. The pharma giant Novartis has licensed one of Monte Rosa’s VAV1 degraders, MRT-6160, which is in Phase 1 clinical trials, in a deal worth up to $2.1 billion.

Monte Rosa also demonstrated that its AI/machine learning platform, QuEEN, can predict which target proteins are compatible with cereblon. The company says its platform can make this determination based on the target protein’s surface. This is possible because while proteins are flexible, they are not infinitely so.

“Our glues really do introduce very specific changes as well,” Castle says. “It is specific changes that we can model and then predict with the algorithms.”