Scripps Research scientists and colleagues show how drugs that eliminate certain disease-driving proteins can be discovered systematically rather than by chance.
February 27, 2026
LA JOLLA, CA—Though many drugs are designed to block the activity of proteins involved in disease, some compounds, called degraders, can go a step further by removing problematic proteins from a cell altogether. “Molecular glue” degraders do this by binding to a target protein and then forcing it to interact with the cell’s natural disposal system. But while molecular glues hold promise for treating diseases driven by hard-to-target proteins, most have been discovered by accident, leaving scientists without a clear roadmap for finding new ones.
Now, a new collaborative study from Scripps Research reports a way to change that. Published in Nature Chemical Biology on February 16, 2026, the study was led by Michael Erb, an associate professor of chemistry at Scripps Research, and Georg Winter, a former principal investigator at the CeMM Research Center for Molecular Medicine of the Austrian Academy of Sciences who’s now the scientific director of the AITHYRA Research Institute for Biomedical Artificial Intelligence of the Austrian Academy of Sciences. Their findings describe a strategy for discovering molecular glues deliberately rather than by chance. The approach starts with small molecules that are already known to “stick” to a specific protein but don’t mark it for removal. By making subtle chemical changes and testing the results, the researchers asked whether those same molecules could be transformed into molecular glues that cause a cell to eliminate a target protein.
“Discovering molecular glues purposely and prospectively has been a long-standing challenge with very few solutions,” says Erb. “What we show here is that it’s possible to take a molecule that already binds to a specific protein and convert it into a molecular glue that eliminates the protein.” In earlier work, Erb and others had identified small “starter” molecules that stick to proteins with key roles in cancer and gene regulation but don’t trigger their removal from a cell. The question was whether subtle chemical changes could give those same molecules entirely new behaviors, ultimately removing a target protein rather than just sticking to it. To find out, the team of the current study generated roughly 3,000 slightly altered versions of each starting molecule. Each version differed by a small chemical add-on, but all were based on the same core structure. The researchers tested these variants in cells to determine whether any caused the target protein to be pulled into contact with the cell’s protein-disposal system.
“Going into our project, one of the biggest unknowns was scale since we didn’t know whether we’d need to test a few hundred molecules or many thousands to find molecular glues,” recalls Erb. “When we started to see signs that some molecules were beginning to trigger protein removal, that’s when it became clear this strategy may actually be viable.”
Using this approach, the team ultimately focused on three compounds that triggered targeted protein degradation. Two of these compounds acted through glue-like mechanisms: binding to the target protein first and then recruiting the cell’s protein-degradation machinery. One compound selectively eliminated the cancer-related eleven-nineteen leukemia (ENL) protein in leukemia cells, while the other induced degradation of BRD4 (a protein that controls gene expression related to cancer) by engaging a component of the degradation machinery that hadn’t previously been harnessed for this purpose.
Importantly, the study also revealed that not all molecular glues work in the same way. In the case of ENL, the newly discovered compound only engaged the degradation machinery after first sticking to the target protein, forming a combined surface that stabilized the interaction. This cooperative mechanism helps explain why small structural changes can have such significant effects, and why these compounds are difficult to predict without testing thousands of chemical variations. Beyond cell-based experiments, the researchers showed that one of the ENL-targeting compounds could reduce levels of the problematic protein in mouse models, suggesting it may be relevant as a starting point for future drug development. Still, Erb is careful to point out what the study doesn’t yet solve.
“This approach starts with a protein that you can already bind with a small molecule, but it doesn’t suddenly make it possible to find drugs for hard-to-target proteins that don’t readily bind to small molecules,” he explains. “Instead, it gives us a way to turn molecules that don’t do much on their own into something more functional.” Erb cites his previous work on FOXA1, which is linked to hormone-driven breast and prostate cancer, as an example. Working with the lab of Benjamin Cravatt, the Gilula Chair in Biology and a professor of chemistry at Scripps Research, Erb’s lab showed that small molecules could bind to FOXA1. The new method now offers a potential way to go further, turning these molecules into compounds that actively control the protein’s fate in a cell. “A lot of progress has been made in finding molecules that bind to proteins that were historically difficult to target with drugs,” Erb points out. “The challenge in many cases is figuring out what to do with those binders, and this novel approach gives us a new way forward.”
In addition to Erb and Winter, authors of the study, “High-throughput ligand diversification to discover chemical inducers of proximity,” include James B. Shaum, Erica A. Steen, Jordan Janowski, Eric M. Bilotta, Paige A. Barta, Natalia Milosevich, Lauren M. Hargis, Timothy R. Bishop, Trever R. Carter, Bryce da Camara, Wen-Ji He, Benjamin F. Cravatt, Keary M. Engle and Bruno Melillo of Scripps Research; Miquel Muñoz i Ordoño, Matthias Hinterndorfer and Fabian Offensperger of the CeMM Research Center for Molecular Medicine of the Austrian Academy of Sciences; Daniela V. Wenge, Hakyung Cheong, Moritz Hunkeler, Abby M. Thornhill, Sydney R. Schweber, Charlie Hatton, Yanhe Wen, Katherine A. Donovan, Scott A. Armstrong and Eric S. Fischer of the Dana-Farber Cancer Institute; Zoe J. Rutter, Hirotake Furihata and Alessio Ciulli of the University of Dundee; and Lucas Dada and Seiya Kitamura of the Albert Einstein College of Medicine.
This work was supported by funding from the Austrian Academy of Sciences; the Austrian Science Fund (projects P7909, P36746 and P5918723); the Baxter Foundation Young Investigator Award; the Boehringer Ingelheim Fonds; the Boehringer Ingelheim Stiftung; Curing Kids Cancer; the European Research Council (under the European Union’s Horizon 2020 research and innovation program; grant agreement 851478); the German Research Foundation (Deutsche Forschungsgemeinschaft; project 511811315); the Innovative Medicines Initiative 2 (IMI2) Joint Undertaking (under grant agreement 875510 [EUbOPEN project]; the IMI2 Joint Undertaking receives support from the European Union’s Horizon 2020 Research and Innovation Program, the European Federation of Pharmaceutical Industries and Associations companies and associated partners KTH Royal Institute of Technology, the Ontario Institute for Cancer Research, Diamond Therapeutics and McGill University); the Japan Society for the Promotion of Science (postdoctoral fellowship 23KJ1669); the National Cancer Institute (grants R01CA280720 and R01CA214608); the National Institutes of Health (grants CA066996, CA259273, K99GM138758 and R35GM155249); the NIH Institutional National Research Service Award (grant T32TR004396); the NIH Office of the Director (grant DP5-OD26380); the Ono Pharma Foundation Breakthrough Science Initiative Awards Program; the United States National Science Foundation (grant CHE-2046286); the V Foundation for Cancer Research; and the Vienna Science and Technology Fund (project LS21-015).
For more information, contact press@scripps.edu See More News