State-of-the-art X-ray Crystallographic Fragment Screening at the Diamond Light Source’s XChem Facility

Writer: Harrison Whitman

Date: Fall 2018

Citation: Whitman, H. (2018). State-of-the-art X-ray Crystallographic Fragment Screening at the Diamond Light Source’s XChem Facility. Rutgers Research Review, 3(1).


My name is Harrison Whitman, and I am a rising sophomore in the School of Arts and Sciences. I am majoring in Chemistry and minoring in Psychology. As part of the Aresty Center’s Summer Science program, I worked in the laboratory of Dr. Eddy Arnold in the Department of Chemistry and Chemical Biology, at the Center for Advanced Biotechnology and Medicine (CABM).

The Arnold laboratory studies the structural biology of human immunodeficiency virus (HIV), the virus that causes acquired immune deficiency syndrome (AIDS). Understanding the structure of biological molecules is important for several reasons. For large molecules such as proteins and nucleic acids, structure is inextricably linked to function. Structural information is valuable because it allows us to rationally design drugs to specifically target a protein or enzyme that plays a role in the etiology of a disease. Structural studies have been major contributors to our collective understanding of HIV and have led to the development of several drugs that have significantly improved therapeutic outcomes and quality of life for over 17 million HIV positive individuals (World Health Organization, 2016). For one example, X-ray crystallographic information collected by the Arnold group advanced the hypothesis that conformationally flexible drugs that could “wiggle” and “jiggle” in three dimensions were more able to overcome resistance mutations in a target protein -- this idea was the guiding principle behind the iteration and eventual development of the anti-HIV drug etravirine, discovered by the Janssen Research Foundation (Das et al, 2004).

A highly effective strategy for structure-based drug design is known as fragment screening. Fragment screening, or as it is more formally known, fragment-based drug discovery, is a method by which libraries of small molecules are evaluated for their binding to a target protein. Fragments that bind to the target serve as starting points for developing drugs. Medicinal chemistry work can be undertaken to synthesize derivatives of a fragment capable of binding to the target (Arnold, 2014). Initial screening using fragments allows for wide swaths of “chemical space”, or permutations of molecular structures, to be sampled in a more time- and cost-effective fashion than with traditional high-throughput screening (Bauman et al, 2013). The binding of these fragments can be detected using techniques such as nuclear magnetic resonance (NMR), surface plasmon resonance (SPR), or, as the Arnold lab principally uses, X-ray crystallography (Arnold, 2014).

X-ray crystallography is a method for determining the arrangement of atoms in a molecule, and is useful for solving the structures of biological macromolecules such as proteins. A fragment that binds to a target protein will create an observable difference in protein structure. By comparing the information obtained from crystals that were soaked with potentially-binding fragments to those without, the binding of that fragment can be deduced (Bauman et al, 2013). The method works by bombarding a target crystal with X-ray radiation. The X-rays are diffracted by the planes of atoms in the crystal, and thus the arrangement of the diffraction pattern depends on the arrangement of those atoms in three-dimensional space (Pauling, 1988).

Though X-ray crystallography has been an indispensable tool for understanding chemical structures for the past century, it is only recently that improvements in beamline technology, automated crystal mounting, data collection, and data processing and analysis have made large-scale fragment screening campaigns possible (Patel et al, 2014). Still, it may prove difficult for groups operating independently, or with limited resources, to assemble a large library of compounds to use as fragments. For example, the Arnold group’s 2013 fragment screening campaign against reverse transcriptase-rilpivirine (an HIV drug targeting reverse transcriptase) complexes relied on both purchased compounds and those gifted by collaborators (Bauman et al, 2013). Owing to the growing popularity and utility of X-ray crystallographic fragment screening, efforts are being made to make the technique more efficient and accessible.

In Oxfordshire, United Kingdom, the Diamond Light Source synchrotron facility has developed a cutting-edge, dedicated fragment screening resource known as XChem, which the Arnold laboratory has been approved to use. The XChem facility streamlines and automates the process from soaking crystals to analyzing diffraction data, “allowing up to 1000 compounds to be screened individually in less than a week” (Diamond Light Source, 2018). In addition to simply accelerating the process, the XChem facility provides several other advantages. Among these are:

Singleton fragment soaking.For most of the history of fragment screening, the relatively slow rate of X-ray diffraction data collection prohibited the screening of individual fragments against individual crystals, one at a time. Instead, crystals were screened against cocktails of several fragments at once, and hits were subsequently verified by individual screens (Bauman et al, 2013). The advancements in the Diamond system make singleton fragment soaking feasible, allowing a higher concentration of each compound to be individually tested against the target crystal (Diamond Light Source, 2018).

Pan-Dataset Density Analysis (PanDDA) interface.The PanDDA algorithm, as developed by Pearce et al (2017), is a powerful method for analyzing X-ray crystallographic data. Recall that the binding of fragments is identified by comparing the structure of proteins soaked with a fragment to the structure of a protein not soaked with a fragment. These changes in structure may be so weak they get lost in the “noise” and fluctuations of measured data. It is also crucial to avoid mistaking solvent molecules binding to the target for fragment compound binding. Thus, a “blank” data set for comparison purposes is essential (Bauman et al, 2013). The PanDDA method analyzes many data sets of protein unbound to any fragment to form a composite blank data set and quantitatively determine when a particular dataset deviates from the standard “ground state” for the crystal form being studied, thereby obviating both of these problems (Pearce et al, 2017). The XChem facility provides an interface to use the PanDDA algorithm, which, coupled with the large number of datasets collected, can help pinpoint changes to the crystal structure of target proteins that may have otherwise gone undetected.

Though the use of the XChem facility represents a great opportunity for the Arnold laboratory, there are still some challenges that must be overcome. For instance, the XChem facility is optimized for using dimethyl sulfoxide (DMSO) as the solvent for fragments, which is known to hinder X-ray diffraction or destabilize protein crystals entirely (Diamond Light Source, 2018; Patel et al, 2014). Another potential roadblock is temperature (Patel et al, 2014). At our lab in CABM, we can grow crystals and soak them with fragments at a variety of temperatures, though there may not be as much control at XChem, which could reduce the quality of our results. To overcome these challenges, we are experimenting with ethylene glycol as an alternative solvent, which XChem has some limited support for, and working to optimize conditions to grow crystals at 20°C, which is also recommended by XChem (Diamond Light Source, 2018). The advances at the XChem facility represent the state-of-the-art in X-ray crystallography and fragment screening, and it will undoubtedly be exciting to see what results we can uncover in the next year by harnessing them.

References

  1. Arnold, E. (2014). “Fragment screening for drug discovery: Efficient approaches for exploring chemical space”. Progress in Biophysics and Molecular Biology, 116, 81.
  2. Bauman, J. D., Patel, D., Dharia, C., Fromer, M. W., Ahmed, S., Frenkel, Y., … and Arnold, E. (2013). “Detecting allosteric sites of HIV-1 reverse transcriptase by X-ray crystallographic fragment screening.” Journal of Medicinal Chemistry, 56(7), 2738–2746.
  3. Das, K., Clark, D. A., Lewi, P. J., Heeres, J., de Jonge, M. R., Koymans, L. M. H., … and Arnold, E. (2004). “Roles of conformational and positional adaptability in structure-based design of TMC125-4165335 (etravirine) and related non-nucleoside reverse transcriptase inhibitors that are highly potent and effective against wild-type and drug-resistant HIV-1 variants.” Journal of Medicinal Chemistry, 47(10), 2550-2560.
  4. Diamond Light Source (2018). Fragment Screening - XChem. Retrieved from https://www.diamond.ac.uk/Instruments/Mx/Fragment-Screening.html.
  5. Patel, D., Bauman, J. D., Arnold, E. (2014). “Advantages of crystallographic fragment screening: Functional and mechanistic insights from a powerful platform for efficient drug discovery.” Progress in Biophysics and Molecular Biology, 116, 92–100.
  6. Pauling, L. (1988). General chemistry. New York, NY: Dover.
  7. Pearce, N. M., Krojer, T., Bradley, A. R., Collins, P., Nowak, P. R., Talon, R. … and von Deflt, F. (2017). “A multi-crystal method for extracting obscured crystallographic states from conventionally uninterpretable electron density.” Nature Communications, 8(15123), 1–8.
  8. World Health Organization (2016). 17 million people with access to antiretroviral therapy. Retrieved from https://www.who.int/hiv/mediacentre/news/global-aids-update-2016-news/en/