The recent discovery of a hidden drug pocket in the cancer protein PKMYT1 by researchers at the Icahn School of Medicine at Mount Sinai has sparked excitement and raised questions about the future of cancer treatment and the role of AI in drug discovery. This groundbreaking finding not only opens up new possibilities for developing more precise cancer drugs but also highlights the limitations of current AI tools in predicting protein structures.
The study, published in the Journal of the American Chemical Society, focused on PKMYT1, a kinase protein that plays a crucial role in cell growth and division. While many kinase inhibitors target the ATP-binding site, which is a common region for drug binding, the researchers discovered a previously unknown binding pocket in PKMYT1 using a combination of AI-based protein prediction tools and laboratory experiments.
This hidden pocket, which was missed by state-of-the-art AI systems, could potentially lead to the development of more selective cancer drugs. The findings suggest that proteins like PKMYT1 are far more flexible than previously thought, constantly shifting between different shapes. This flexibility, combined with the ability of small chemical changes to alter binding sites, highlights the dynamic nature of protein-drug interactions.
Co-senior and co-corresponding author Avner Schlessinger emphasizes the importance of this discovery, stating, 'Our study shows both the power and the limitations of AI in drug discovery. AI was very accurate when predicting known protein shapes, but it missed a completely unexpected binding pocket that we could only uncover experimentally.' This underscores the need for experimental validation in drug discovery, even with the advancements in AI.
The research team's use of AI tools like AlphaFold2 and AlphaFold3, along with molecular dynamics simulations, demonstrated that current computational approaches may not fully capture the complexity of protein structures. The discovery of the hidden pocket and its sensitivity to small chemical modifications also raises questions about the dynamic nature of protein binding and the potential for developing more selective drugs.
Co-senior and co-corresponding author Michael Lazarus notes, 'One of the most surprising findings was that a very small chemical modification caused the molecule to switch from binding in this hidden pocket to binding in a much more conventional way. That tells us these proteins are incredibly dynamic and sensitive to subtle molecular changes.'
The implications of this research are far-reaching. It suggests that future AI systems could benefit from improved methods to predict hidden and dynamic protein states. Additionally, the discovery of the hidden pocket provides a new avenue for developing more selective cancer drugs, potentially reducing the toxicity and specificity challenges associated with traditional kinase inhibitors.
The investigators plan to further optimize and test the compounds identified in the study, aiming to develop more potent drugs that target the newly discovered site. They also hope to explore the existence of similar hidden pockets in other cancer-related kinases. This research not only advances our understanding of protein-drug interactions but also paves the way for the development of more effective and targeted cancer therapies.
