In this new blog post in our series on structural biology and drug design, we want to bring the attention of our readers to the revival of room-temperature (RT) protein crystallography using synchrotron radiation. After more than two decades of cryo-crystallography, finding someone with skills in mounting protein crystals for RT-crystallography won't be easy. It would be safe to say that most of the young crystallographers today haven’t even seen an X-ray capillary of the type used in RT-crystallography.

Robert Steiner has put together a virtual issue of Acta Cryst. D, Structural Biology, with articles on RT biological crystallography. The issue contains a collection of papers highlighting some topics and emerging trends in the new incarnation of RT protein crystallography.

Why RT-crystallography?

An obvious advantage of doing RT-crystallography is, as stated at the site of MiTeGen, a company that offers products and tools for carrying out X-ray crystallography:” Roughly 98% of low-temperature data sets collected on structural biology beamlines do not yield diffraction data of sufficient quality to determine a structure. Room temperature measurements are essential to determine whether the cause of poor diffraction is poor as-grown crystal quality, sample damage caused by ligand and/or cryoprotectant soaks, and/or sample damage caused by the flash cooling process itself. Far too much effort is wasted searching for cryoprotection conditions for crystals that are poorly ordered to start with."

Other shortcomings of cryo-crystallography are listed in a review article by Robert E. Thorn in the above-mentioned issue:

- Regions of ordered density in a cryo-structure may be disordered at RT (and vice versa). Disorder at RT may be of functional relevance. However, this is missed in cryo-crystallography since most structures are refined to a single conformation. RT-data collection can facilitate the identification of functionally important alternative conformations at active sites and elsewhere.

- The cryoprotectant used in cryo-crystallography may perturb side chain conformations. It may also contribute additional electron density, including in active site regions.

- Cryo-cooling may degrade crystal lattice order, even if thermal disorder is reduced. It may also introduce crystal non-isomorphism and high mosaicity. Variability in cryoprotectant soaks and cooling may also result in non-isomorphism.

These realizations and the recent technical developments discussed in the review by Robert E. Thorn triggered the revival of RT-crystallography.

- the development of X-ray-free electron laser (XFEL) sources and serial crystallography at synchrotrons

- development of serial-sample delivery and software for processing and modeling data collected from a large number of small crystals.

- improvements in detector speed and sensitivity; new techniques for timeresolved study of reactions and conformational dynamics within crystals.

Conformational diversity and ligand binding

These new findings show that RT and multi-temperature crystallography can be used to study functionally important conformations. They can also be used to find temperature-dependent differences in protein-ligand interactions, such as new binding sites and unique binding poses. Such studies may provide a more complete understanding of protein-ligand interactions, making RT crystal structures a better representation of the functionally relevant conformation of the protein and, thus, more efficient for use in drug discovery.

An example of a study of ligand-binding discrepancies between cryo-cooled and RT-temperature structures is in the work by Huang and co-workers who used a temperature-variation approach to examine the binding of TL00150, a 175.15 Da endothiapepsin inhibitor. The authors report the details of the experimental setup for the temperature-dependent measurements and show that at 298 K, depending on DMSO concentration, the fragment TL00150 had two binding positions. In contrast, at cryo-temperature (100 K), it was only observed in a single position at the S1 part of the active site pocket. Subsequently, to assess ligand occupancy at different temperatures, data for a series of structures with bound ligand were collected. The data showed different ligand binding states with different occupancies at the S1 and S1' sites. This was explained by a flexible loop close to the protein's active site.

The new technical developments, by making RT and multi-temperature measurements relatively easily accessible at synchrotrons, provide new opportunities for studying functionally important conformations that may be relevant for drug design. And last, but not least, we could actually start testing crystals directly at synchrotrons at room temperature without all the hassle of finding a suitable cryoprotectant and freezing!