Diffuse X-ray scattering from correlated motions in a protein crystal

Abstract

Protein dynamics are integral to biological function, yet few techniques are sensitive to collective atomic motions. A long-standing goal of X-ray crystallography has been to combine structural information from Bragg diffraction with dynamic information contained in the diffuse scattering background. However, the origin of macromolecular diffuse scattering has been poorly understood, limiting its applicability. We present a finely sampled diffuse scattering map from triclinic lysozyme with unprecedented accuracy and detail, clearly resolving both the inter-and intramolecular correlations. These correlations are studied theoretically using both all-atom molecular dynamics and simple vibrational models. Although lattice dynamics reproduce most of the diffuse pattern, protein internal dynamics, which include hinge-bending motions, are needed to explain the short-ranged correlations revealed by Patterson analysis. These insights lay the groundwork for animating crystal structures with biochemically relevant motions.

Reproducibility of protein x-ray diffuse scattering and potential utility for modeling atomic displacement parameters

“…In step (4), pixel positions were corrected by the parallax broadening effect in the PILATUS 6M detector, 20 and raw pixel intensities were converted to scattering intensities by applying polarization, 16 solid-angle, 16 and detector absorption corrections. 21 …”

“…In previous work, three different methods were mentioned regarding removal of Bragg pixels, by either filtering out Bragg pixels, 16–18 replacing intensities, 34 or preserving Bragg peak intensities together with diffuse scattering features. 21 In this work, we chose to filter out all pixels in Bragg peak positions as we were interested in large-scale diffuse features that vary on a length scale longer than the separation between Bragg peaks. The other two methods are useful for obtaining more finely sampled datasets and analyzing sharper diffuse scattering features.…”

“… 7–15 Since the development of modern diffuse data processing methods, 16,17 protein diffuse scattering studies have mostly focused on working with three-dimensional (3D) datasets. In addition to improvements in light sources and detectors, notable developments in 3D data processing include finer sampling in reciprocal space to model long-range correlations, 18 rescuing useful diffuse data from experiments designed for Bragg diffraction, 19 extracting finely sampled 3D datasets from serial femtosecond x-ray crystallography (SFX) experiments with x-ray free-electron lasers (XFELs), 4 increasing data quality via improved rejection of the solvent contribution and multivariate analysis methods, 20 and a major advance in the scaling and merging of data from multiple crystals, 21 yielding a substantial improvement in data quality.…”

Combining X-rays, neutrons and electrons, and NMR, for precision and accuracy in structure–function studies

“…Where this is true these are missed opportunities to connect the structures and their dynamics. For overviews, and descriptions of the benefits of analysing the diffuse scattering, see Wall et al (2018), Meisburger et al (2020, and for a detailed example see de Klijn et al (2019).…”

Advances in methods for atomic resolution macromolecular structure determination

“…Additionally, computational modelling software that can use experimental X-ray diffraction data to create models of alternative conformations or entire conformational ensembles is now available 132 – 136 , allowing crystallographers to maximize the amount of structural and biophysical information that can be accessed from data collected at non-cryogenic temperatures. Frontiers in this area of research include the measurement of diffuse (non-Bragg) scattering for the analysis of protein dynamics 137 – 141 and the incorporation of full molecular dynamics simulations (of crystallized molecules) into the modelling and refinement procedure 142 – 145 .…”