Baylor Team Develops COOKIE-Pro to Optimize Potency and Selectivity of Covalent Drugs

Researchers at Baylor College of Medicine have created a data analysis method called COOKIE-Pro (Covalent Occupancy Kinetic Enrichment via Proteomics) to better understand how covalent inhibitors interact with proteins inside cells. This approach, described in a Nature Communications paper, enables scientists to measure both how strongly and how quickly these drugs bind to thousands of potential protein targets at once, providing a more complete picture than past techniques.

Covalent inhibitors work by forming permanent bonds with target proteins. This makes them highly effective. However, the same property can lead to unintended interactions with other proteins, sometimes resulting in side effects. “Covalent inhibitors, which include well-known drugs like aspirin and the cancer therapeutic ibrutinib, are highly effective because they form a strong, permanent bond with their target protein,” explained Jin Wang, corresponding author of the study. “However, this strength can be a double-edged sword; these drugs can also bind to unintended off-target proteins, potentially leading to unwanted side effects.”

Optimizing these compounds requires balancing affinity, or how strongly they are drawn to a target, with reactivity, or how quickly they form a bond. According to first author Hanfeng Lin, existing methods made it difficult to measure these properties across many proteins. “COOKIE-Pro gives us a comprehensive map in which we can see for the first time, not just if a drug binds off-target, but how well and how fast, which is critical information for drug designers,” Lin said.

COOKIE-Pro works through a two-step system. Scientists first expose broken-down cells to a covalent drug so it can attach to its targets. They then add a specially designed probe that binds to any remaining unoccupied protein sites. By measuring the probe’s binding with mass spectrometry, researchers can calculate how much each protein was targeted by the drug, revealing both binding affinity and inactivation speed.

The team validated the method using spebrutinib and ibrutinib. Results confirmed known data but also uncovered new findings, such as spebrutinib binding more than 10 times more strongly to an off-target protein, TEC kinase, compared to its intended target, BTK. For ibrutinib, COOKIE-Pro successfully mapped its recognized off-target interactions. 

Beyond validation, the team showcased COOKIE-Pro’s potential for large-scale screening. A streamlined version applied to 16 drug fragments generated thousands of protein interaction profiles, highlighting its use in the earliest phases of drug discovery. 

“The ultimate goal is rational drug design,” Wang said. “A drug might appear potent because it binds quickly, but if that is simply because it has a ‘hot’ reactive group, it might cause side effects by binding everywhere. COOKIE-Pro allows us to separate that intrinsic reactivity from true binding affinity. We can now help chemists prioritize compounds that are potent because they bind specifically to the right target, not just because they are broadly reactive. This is a crucial step toward creating the next generation of highly selective and safer covalent medicines.”