Targeting RNA with Small Molecules using State-of-the-Art Methods Provides Highly Predictive Affinities of Riboswitch Inhibitors

Krystel El Hage,Louis Lagardère,Narjes Ansari,Chengwen Liu,Florent Hedin,Jérôme Hénin,Jay W. Ponder,Pengyu Ren,Jean-Philip Piquemal

DOI: https://doi.org/10.26434/chemrxiv-2024-s2lqq

2024-10-17

Abstract:Targeting RNA with small molecules represents a promising yet relatively unexplored avenue for the design of new drugs. Nevertheless, challenges arise from the lack of computational models and techniques able to accurately model RNA systems, and predict their binding affinities to small molecules. Here, we tackle these difficulties by developing a tailored state-of-the-art approach for absolute binding free energy calculations of RNA-binding small molecules. To do so, we combine the advanced AMOEBA polarizable force field, which accounts for both accurate multipolar electrostatics and many-body effects, to the newly developed Lambda-Adaptive Biasing Force (Lambda-ABF) scheme associated to refined restraints allowing for efficient sampling. Furthermore, to capture the free energy barrier associated to challenging RNA conformational changes, we apply machine learning to identify effective collective variables in order to use them into further enhanced sampling simulations based on an evolution of metadynamics. Applying this computational protocol to a complex Riboswitch-like RNA target, we demonstrate quantitative predictions. These results pave the way for the routine application of free-energy simulations in RNA-targeted drug discovery, thus providing a significant reduction in their failure rate.

Chemistry

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to develop an advanced computational method for accurately predicting the binding affinity between small molecules and RNA. Specifically, the authors focus on several key challenges in RNA - targeted drug design:

1. **Accuracy of computational models**: The special properties of RNA systems, such as being highly dynamic, charged, and having interactions with ions and polarized water molecules, make it difficult for existing computational models to accurately simulate RNA systems and their binding affinities with small molecules.

2. **Challenges in free - energy calculations**: The calculation of absolute binding free energy (ABFE) is more challenging than that of relative binding free energy (RBFE) because it does not rely on reference compounds and is not limited to ligands with similar binding patterns. Especially for complex biomolecules like RNA, calculating its binding affinity with small molecules is a huge challenge.

3. **Impact of conformational changes**: RNA may undergo significant conformational changes before and after binding to small molecules, and such changes pose additional challenges to free - energy calculations.

To address these challenges, the authors developed a new computational protocol that combines the AMOEBA polarizable force field and the Lambda - Adaptive Biasing Force (Lambda - ABF) method to achieve efficient sampling and accurate prediction. In addition, they also use machine - learning techniques to identify effective collective variables for use in enhanced - sampling simulations, thereby more accurately capturing the free energy of RNA conformational changes.

Through these methods, the authors hope to provide a reliable computational tool for RNA - targeted drug design, thereby reducing the failure rate in the drug discovery process.