Protein crystal structure determination with the crystallophore, a nucleating and phasing agent

Abstract

Obtaining crystals and solving the phase problem remain major hurdles encountered by bio-crystallographers in their race to obtain new high-quality structures. Both issues can be overcome by the crystallophore, Tb-Xo4, a lanthanide-based molecular complex with unique nucleating and phasing properties. This article presents examples of new crystallization conditions induced by the presence of Tb-Xo4. These new crystalline forms bypass crystal defects often encountered by crystallographers, such as low-resolution diffracting samples or crystals with twinning. Thanks to Tb-Xo4's high phasing power, the structure determination process is greatly facilitated and can be extended to serial crystallography approaches.

Keywords: Tb-Xo4; anomalous-scattering-based methods; de novo phasing; macromolecular crystallization; protein crystallography; serial crystallography; the crystallophore.

Figure 1

P. horikoshii protease 1 phasing based on data collected by the MeshAndCollect approach. (a) Dendrogram resulting from HCA on 51 sub-data sets with clustering according to correlation coefficients. The selected cluster (identified with a blue dashed rectangle) comprises 27 data sets (threshold of 0.3). (b) Anomalous Patterson map (Harker section w = 0.25). (c) Experimental electron density resulting from phasing with the CRANK2 pipeline (contoured at 1σ). (d) Automatically built model of PhP1.

Figure 2

pb9 phasing based on data collected through the MeshAndCollect approach. (a) Dendrogram resulting from HCA on 55 sub-data sets with clustering according to unit-cell variation. The selected cluster (identified with a blue dashed rectangle) comprises 32 data sets (threshold of 0.8). (b) Anomalous Patterson map (Harker section v = 0.5). (c) Experimental electron density resulting from phasing with the CRANK2 pipeline (contoured at 1σ). (d) Automatically built model of pb9.

Figure 3

Samples obtained from M. thermolithotrophicus by native purification. (a) SDS-PAGE gels of the different purified protein fractions. (b) Details of the results of crystallization screening performed on the four protein fractions. The number of unique crystallization hits is depicted in grey for the native protein without Tb-Xo4 and with dots for the protein supplemented with 10 mM Tb–Xo4. The conditions where crystals were obtained irrespective of the presence of Tb-Xo4 are represented in grey with dots. Owing to sample quantity availability, fractions A and D were only evaluated in the presence of Tb-Xo4. (c)−(f) Examples of crystals resulting from fractions A–D, respectively. Crystallization was performed in the presence of 10 mM Tb-Xo4.

Figure 4

Experimental electron density map (contoured at 1σ) resulting from the SAD phasing of the data collected (a) at the terbium L _III-absorption edge on crystal form 1 and (b) at a wavelength of 0.977 Å on crystal form 2 of the FprA protein.

Figure 5

(a) Two orthogonal views of the experimental electron density map (contoured at 1σ) resulting from the SAD phasing of the ThiS crystal (form 1) soaked in 100 mM crystallophore. Cartoon representations of the ThiS biological unit are shown. (b) Anomalous Fourier synthesis (in red contoured at 8σ) computed with data collected on ThiS crystal form 3.

Figure 6

AdkA protein. (a) Experimental electron density map (contoured at 1σ) resulting from the ab initio phasing based on data collected on a crystal soaked in 50 mM crystallophore. (b) Portions of the model built automatically, leading to the identification of the AdkA protein sequence superimposed on the electron density map depicted in (a).