High-Confidence Placement of Fragments into Electron Density Using Anomalous Diffraction-A Case Study Using Hits Targeting SARS-CoV-2 Non-Structural Protein 1.

The identification of multiple simultaneous orientations of small molecule inhibitors binding to a protein target is a common challenge. It has recently been reported that the conformational heterogeneity of ligands is widely underreported in the Protein Data Bank, which is likely to impede optimal exploitation to improve affinity of these ligands. Significantly less is even known about multiple binding orientations for fragments (<300 Da), although this information would be essential for subsequent fragment optimisation using growing, linking or merging and rational structure-based design. Here, we use recently reported fragment hits for the SARS-CoV-2 non-structural protein 1 (nsp1) N-terminal domain to propose a general procedure for unambiguously identifying binding orientations of 2-dimensional fragments containing either sulphur or chloro substituents within the wavelength range of most tunable beamlines. By measuring datasets at two energies, using a tunable beamline operating in vacuum and optimised for data collection at very low X-ray energies, we show that the anomalous signal can be used to identify multiple orientations in small fragments containing sulphur and/or chloro substituents or to verify recently reported conformations. Although in this specific case we identified the positions of sulphur and chlorine in fragments bound to their protein target, we are confident that this work can be further expanded to additional atoms or ions which often occur in fragments. Finally, our improvements in the understanding of binding orientations will also serve to improve the rational optimisation of SARS-CoV-2 nsp1 fragment hits.

FEM: feature-enhanced map.

A method is presented that modifies a 2mFobs - DFmodel σA-weighted map such that the resulting map can strengthen a weak signal, if present, and can reduce model bias and noise. The method consists of first randomizing the starting map and filling in missing reflections using multiple methods. This is followed by restricting the map to regions with convincing density and the application of sharpening. The final map is then created by combining a series of histogram-equalized intermediate maps. In the test cases shown, the maps produced in this way are found to have increased interpretability and decreased model bias compared with the starting 2mFobs - DFmodel σA-weighted map.