The impact of synchrotron radiation on structure-based drug design

In this article, we highlight an interesting study conducted by the AstraZeneca Gothenburg structural biology team on the impact of synchrotron radiation in drug design (Käck & Sjögren, 2024). Using the team’s internal repository, which contains 3717 unique structures delivered over 20 years (2004-2023) to 186 non-oncology therapy area projects, the authors examine the evolution and current state of synchrotron use in structure-based drug design. It’s not common to see such analyses published by a major pharmaceutical company, and the Gothenburg team has certainly done an excellent job here. Astra AB, which became AstraZeneca in 1999, has a strong tradition of using X-ray crystallography in drug design. This likely reflects Sweden’s prominent position in the field, with Uppsala being one of the world’s leading centers for protein crystallography as early as the 1970s. Currently, the company has dedicated crystallography teams in Gothenburg and Cambridge, which, according to the authors, deliver around 800 unique protein–ligand complex X-ray structures per year, in addition to a group focused on cryo-EM.

The “synchrotron-only” approach in acquiring X-ray data
The first revelation by the analysis was that a significant shift over the past 20 years has been the company’s transition from a combined in-house and synchrotron data collection model to a ‘synchrotron-only’ approach, which occurred between 2018 and 2019. They note that the change was made possible by several factors, including easier access to synchrotrons, the ability to collect data remotely, and the development of high-throughput crystallography and streamlined workflows at synchrotrons, all of which facilitate the use of synchrotron radiation in drug design. Furthermore, advances in synchrotron technology, such as more stable and focused beams, faster detectors, reliable sample changers, and automated crystal characterization and data collection, have enabled the complete acquisition of a dataset within minutes, thereby removing data collection as a bottleneck in structure determination.

These developments have transformed the way X-ray diffraction data are collected, with one consequence being the growing popularity of crystallographic fragment screening (see our blog post on this subject), which needs 100s of datasets from multiple crystals. This also resulted in a significant increase in the total number of datasets collected at synchrotrons since the practice of testing crystals at a home source was abandoned. Another interesting point noted by the authors is that the choice of a synchrotron for data collection mainly depends on practical considerations, with the infrastructure surrounding the sample and data handling being key factors. Another critical and often unpredictable factor is the reliability of shipping dewars with crystals to synchrotrons, which favors synchrotrons that receive reliable dewar deliveries. This is one of the questions we at SARomics Biostructures also sometimes need to address. However, our proximity to the BioMax beamline at the MAX IV synchrotron in Lund ensures a smooth and secure transfer of the crystals to the beamline. The authors also mention the synchrotron in Lund, which is only a few hours’ drive from Gothenburg, as their primary site for data collection.

Impact of structural information across small-molecule drug discovery projects
Another vital aspect of the use of synchrotron radiation in drug design, discussed by the authors, is the timely delivery of crystal structures to projects over time. They point out that access to structural information influences a small-molecule drug discovery project throughout its entire duration, which lasts 4–7 years. This includes initial target validation, selection of a drug design strategy, identification of a method for determining a starting point for drug design, lead generation, and candidate drug selection.

Within the so-called ‘design, make, test, analyze (DMTA)’ framework, where structure-guided hypotheses are generated based on the outcomes from the previous iteration, data from multiple types of screening and assays converge to design a compound that can advance to clinical trials. To maximize the impact of structural information, structures should be delivered promptly to inform current chemical design, indicating that the speed of data generation is a crucial factor throughout the entire process. This assumes that robust and reproducible crystallization conditions have been established and that quick, reliable access to synchrotron facilities is available.

This is the first part of my review of the paper by Käck & Sjögren (2024). In the second part, I will discuss the use of synchrotrons in some projects presented by the authors. Follow us on LinkedIn to make sure that you don’t miss our future posts. For information on our services, please visit our X-ray crystallography services page.