Probing ligand binding of endothiapepsin by `temperature-resolved' macromolecular crystallography

Abstract

Continuous developments in cryogenic X-ray crystallography have provided most of our knowledge of 3D protein structures, which has recently been further augmented by revolutionary advances in cryoEM. However, a single structural conformation identified at cryogenic temperatures may introduce a fictitious structure as a result of cryogenic cooling artefacts, limiting the overview of inherent protein physiological dynamics, which play a critical role in the biological functions of proteins. Here, a room-temperature X-ray crystallographic method using temperature as a trigger to record movie-like structural snapshots has been developed. The method has been used to show how TL00150, a 175.15 Da fragment, undergoes binding-mode changes in endothiapepsin. A surprising fragment-binding discrepancy was observed between the cryo-cooled and physiological temperature structures, and multiple binding poses and their interplay with DMSO were captured. The observations here open up new promising prospects for structure determination and interpretation at physiological temperatures with implications for structure-based drug discovery.

Keywords: conformational heterogeneity; endothiapepsin; fragment binding; protein plasticity; room temperature macromolecular crystallography.

Figure 1

Sample setup and humidity-controlled chamber. (a) Crystal-harvesting loop on a magnetic base with the protection of a plastic capillary. (b) Overview of the humidity-controlled device at PXII-X10SA. (c) The internal details of the humidity-controlled RT box.

Figure 2

Helical-like data collection and timestamp for data sets with the corresponding temperature. (a) An enlarged view of a crystal harvested in a loop with the crystal tip slightly touching the reservoir solution. A rectangular red dashed line outlines the crystal and the bookmarked positions are illustrated in pink with the data-set numbers. (b, c) The timestamp of the data sets and the corresponding temperatures for the temperature ramp-up at 15%(v/v) DMSO and ramp-down at 5%(v/v) DMSO experiments, respectively. The blue circles and red squares represent the N₂ stream temperature and the estimated actual temperature at the sample, respectively. The numbers above and below the x axis indicate the number of the data set and the time, respectively. The 10 min indicated in the figures is the equilibration time for the experiment temperature before data collection. The arrow lines in the figures are indicative and are not drawn on the scale of the real-time line.

Figure 3

Comparison of 2F _o − F _c electron-density maps at 298 K around (a)–(f) the active site and (g)–(l) the flap domain (residues Ser76–Ser86) of EP–TL00150 structures as a function of DMSO concentration. The catalytic dyad, Asp35 and Asp219, and also Asp119, TL00150 and DMSO, are shown as stick representations. The S1 and S1′ positions of the EP active-site pocket and flap domain are shown in (f), (a) and (g), respectively. The 2F _o − F _c electron-density maps contoured at 1σ are shown in orange for TL00150 and in blue for the rest of the residues and DMSO. The concentration trend of DMSO is indicated by a grey triangular bar. Lys110–Ser120 are omitted from the figures for clarity.

Figure 4

Molecular interactions of TL00150 and EP. Molecular interactions of EP–TL00150 in (a) S1′, (b) both S1′ and S1 and (c) S1 of EP. All relevant hydrogen bonds are depicted as dashed lines. The flap domain, the S1 and S1′ positions of the EP active-site pocket and residues involved in the hydrogen-bonding network are shown. The DMSO and water molecules with a blue shadow are not observed in some RT structures. The residues with a red shadow show an alternative conformation in 100 K structures.

Figure 5

Comparison of EP and EP–TL00150 structures at 100 K. (a) EP structure with 10%(v/v) DMSO. (b) TL00150-soaked EP structure with 10%(v/v) DMSO. The r.m.s.d. (root-mean-square deviation) of EP with 10%(v/v) DMSO to TL00150-soaked EP with 10%(v/v) DMSO is 0.178 Å (residues 1–330). EP with 10%(v/v) DMSO and TL00150-soaked EP with 10%(v/v) DMSO are coloured green and grey, respectively. The flap domain, catalytic dyad (Asp35 and Asp219), Asp119, DMSO, water and TL00150 are depicted. The 2F _o − F _c electron-density maps contoured at 1σ are shown in blue for the flap domain, Asp35, Asp119 and Asp219, in cyan for DMSO molecules, in red for water and in orange for TL00150.

Figure 6

Comparison of the electron-density maps of TL00150 and the flap domain in ramp-up and ramp-down experiments. The electron density of TL00150 and surrounding EP residues is shown at (a) 276 K (ramp-up data set 1), (b) 303 K (ramp-up data set 9), (c) 306 K (ramp-up data set 14), (d) 298 K (ramp-down data set 1), (e) 268 K (ramp-down data set 12) and (f) 267 K (ramp-down data set 18). The 2F _o − F _c electron-density maps contoured at 1σ are shown in orange for TL00150 and in blue for the rest of the residues. The temperatures discussed here were measured from the sample using an infrared camera (details are described in Section 2.3).