Unlocking GSK3β's Secrets: How Its Many Shapes Can Guide Better Drug Design

What is it about?

Glycogen synthase kinase 3 beta (GSK3β), a critical protein kinase involved in numerous diseases, has been extensively studied and crystallized hundreds of times. However, a significant challenge for researchers has been the absence of a single, curated resource that consolidates and analyzes this vast structural data. This means anyone wanting to understand GSK3β's conformational landscape has to manually sift through countless publications, PDB entries, and online databases. This manuscript directly addresses this bottleneck by providing the first comprehensive conformational landscape of GSK3β. We systematically classified 111 unique crystal structures and 88 distinct co-crystallized ligands. To achieve this, we extracted data from KinCoRe, which provides classifiers such as active/inactive states and various dihedral angles. Importantly, while KinCoRe offers valuable individual classifications, it does not categorize overall conformational states. Therefore, we manually grouped combinations of KinCoRe classifiers, along with additional criteria we developed, to define eight distinct conformational substates and evaluate their relationships with bound ligands. This work not only resolves a critical knowledge gap for GSK3β but also establishes a systematic framework that can be applied to other protein kinases, with future potential for automated, AI-driven conformational cataloging.

Featured Image

Photo by Willy the Wizard on Unsplash

Photo by Willy the Wizard on Unsplash

Why is it important?

Drug discovery is an incredibly costly and time-consuming endeavor, with a high percentage of lead compounds failing due to issues like lack of selectivity or off-target side effects. This is a major bottleneck, especially in kinase drug discovery. Our aim is to significantly strengthen the early stages of the drug discovery process to help avoid these downstream problems. By meticulously mapping each conformation of GSK3β, our study provides a crucial tool for structure-based drug design. This detailed understanding of how the protein's shape changes when interacting with different compounds can guide the selection of compatible conformations for lead compounds, thereby reducing false positives and negatives in computational screening. Ultimately, this precise conformational knowledge can accelerate the development of more effective, selective, and potent therapies.

Perspectives