Fragment-based screening is now well-established as a powerful approach to early drug, or ‘lead’, discovery. The principle is to identify weakly-binding compounds (‘fragments’) by screening a limited library of compounds, with resulting hits serving as starting points for chemical elaboration to achieve potency. Of the many suitable biophysical techniques, X-ray crystallography was one of the first to be used, and is the most directly informative1; however, the experimental overheads have historically been too high for it to be widely used for primary screening.
The full X-ray screening experiment has now been implemented as a highly streamlined process at the Macromolecular Crystallography beamline I04-1: up to 500 crystals can be soaked and harvested in a day, with data collection taking 24 hours of beamtime (one crystal every three minutes). This is a partnership between Diamond and the University of Oxford, through a joint group led by Frank von Delft that links the I04-1 team with the Protein Crystallography group of Oxford’s Structural Genomics Consortium (SGC).
Patrick Collins, Postdoctoral Research Associate on beamline I04-1, in the fragment screening lab.
Users generate the crystals in their home lab, and then come to Diamond to perform soaking and harvesting in I04-1’s Lab 36 (pictured) with a local contact available for support. The automated X-ray data collection means that once the crystals have been harvested, users do not need to stay on the beamline and can monitor their experiments remotely.
Data analysis builds on the existing automatic data processing, with new tools now developed that robustly detect even very weak hits, streamline density interpretation and refinement, allow analysis and presentation of hit results, and rapid deposition of hit structures to the PDB. Additionally, a new fragment library is being assembledand characterised, to be available for unrestricted use by the end of 2015: it was designed not only to be maximally compatible with X-ray screening but also provide rapid and robust synthetic routes to chemical elaboration. Coupled with a new computational chemistry algorithm that allows prioritising chemical analogues, the ambition is to provide users not only with initial hits but also a set of realistic synthetic strategies that maximise their chances of generating compounds with sufficient potency to justify further research investment. All the above advances are being prepared for publication as well as full deployment at Diamond.
For generating suitable crystals, the group at the University of Oxford has also validated a surface mutation approach proposed by Dr John Hunt2. Support for this approach will be available via Instruct3.
To discuss potential applications for XChem fragment screening process, contact I04-1 Principal Beamline Scientist for the I04-1 beamline, Frank von Delft: firstname.lastname@example.org
Academic access is granted via peer review panel, and proposals can be submitted online. Find out more here: http://www.diamond.ac.uk/Beamlines/Mx/Fragment-Screening/Applying-for-Fragment-Screening.html
Commercial users can contact Alex Dias in the Industrial Liaison Office: email@example.com
Singleton soaking – Our recommended procedure is to soak crystals in one compound, which allows for a higher effective concentration per soak (>200 mM nominal concentration)
Cocktail soaking – We also support libraries prepared by soaking in cocktails, and on-the-fly cocktails where required
Fragment libraries – We have a compound collection ready for use consisting of ~800 compounds from the Maybridge Ro3 core set of 1,000 compounds. Users can also bring their own libraries, provided they are in ECHO-compatible plates
Harvesting – Crystal harvesting is manual but robot-assisted (i.e. not fully automated): this enables fast enough mounting (>100 crystals/hour in favourable cases) that this is no longer a bottleneck
Crystals – Ideal crystals are those that grow easily in many drops, are chunky (rather than needles), don’t stick to the bottom, consistently diffract to high resolution, tolerate high solvent concentrations, and don’t require complicated cryoprotection. Nonideal crystals can certainly be used, the main consequence is that many steps will be slower.
1. Patel, D. et al. Advantages of crystallographic fragment screening: Functional and mechanistic insights from a powerful platform for efficient drug discovery. Prog Biophys Mol Bio.116(1-2), 92-100 (2014). DOI: 10.1016/j.pbiomolbio.2014.08.004
2. Nicholson Price II, W. et al. Understanding the physical properties that control protein crystallization by analysis of large-scale experimental data. Nat Bio. 27, 51-57 (2009). DOI: 10.1038/nbt.1514
3. Instruct: Integrating Biology. http://www.structuralbiology.eu/
Diamond Light Source is the UK's national synchrotron science facility, located at the Harwell Science and Innovation Campus in Oxfordshire.
Copyright © 2017 Diamond Light Source
Diamond Light Source Ltd
Harwell Science & Innovation Campus