Zachary P. Shultz, Ansar Lee-Sam, Yun-Pu Chang, Luxin Sun, Dylan Grassie, Alessio Gabellini, Kyle Pedretty, Thomas Scattolin, Victoria Izumi, Bin Fang, Samer Sansil, Ramu Kakumanu, Lukasz Wojtas, John Koomen, Ernst Schönbrunn, Andrii Monastyrskyi, Derek Duckett, and Justin M. Lopchuk P. Shultz et al.
Science393,408-416(2026).DOI:10.1126/science.adx7219
Structured Abstract
INTRODUCTION
Covalent drugs are transforming targeted therapy by forming durable bonds with disease-driving proteins, yet most rely on a narrow set of reactive groups, particularly acrylamides. These conventional approaches, although effective, can lead to off-target interactions and restrict broader application by their limited structural design. Expanding covalent drug design beyond these established chemotypes is essential to improve both selectivity and therapeutic performance.
RATIONALE
We sought to establish a general platform for replacing acrylamide-based covalent reactive groups in complex drug molecules with alternative chemotypes that offer improved control over reactivity. Our approach centers on bicyclobutanes that are integrated with sulfur-based functional groups commonly used in medicinal chemistry. To enable broad application, we developed a reagent-based strategy that allows these strain-release elements to be installed at the final stage of a synthesis from widely accessible amine precursors. This modular S(IV)-based platform provides a unified entry to multiple sulfur oxidation states and connectivity patterns, enabling systematic tuning of covalent reactivity and target engagement while preserving the parent-drug architecture. By design, this approach allows direct, head-to-head comparison with established covalent inhibitors.
RESULTS
We developed stable, scalable reagents that enable efficient late-stage installation of strain-release bicyclobutane groups across a wide range of clinically relevant scaffolds, including multiple approved kinase inhibitors. This strategy enables direct bioisosteric replacement of acrylamide warheads without modifying the underlying pharmacophore. These S(VI) strain-release groups are highly chemoselective for thiols, and their intrinsic reactivity can be tuned over a broad range through structural modification.
Notably, compounds with similar intrinsic reactivity displayed markedly different levels of target inhibition, demonstrating that productive covalent engagement depends not only on electrophile reactivity but also on molecular orientation within the protein binding site. In cellular systems, the modified inhibitors retained potent activity and effectively suppressed target signaling. Structural analysis confirmed covalent bond formation at the intended site. Across kinase panels and proteome-wide profiling experiments, these strain-release analogs displayed improved selectivity and reduced off-target interactions relative to their acrylamide counterparts. Importantly, these advances translated beyond in vitro systems, with strain-release analogs demonstrating favorable pharmacokinetic properties and efficacy in preclinical in vivo mice models.
CONCLUSION
This work establishes a reagent-enabled, late-stage functionalization platform for the bioisosteric replacement of acrylamides using strain-release bicyclobutanes, bridging chemical innovation to preclinical validation. More broadly, it further supports that effective covalent inhibition is governed not only by intrinsic electrophile reactivity but also by its integration with molecular recognition, providing a framework for designing more selective and clinically effective covalent therapies.
