On the possibility of using polycrystalline material in the development of structure-based generic assays

Abstract

The discovery of ligands that bind specifically to a targeted protein benefits from the development of generic assays for high-throughput screening of a library of chemicals. Protein powder diffraction (PPD) has been proposed as a potential method for use as a structure-based assay for high-throughput screening applications. Building on this effort, powder samples of bound/unbound states of soluble hen-egg white lysozyme precipitated with sodium chloride were compared. The correlation coefficients calculated between the raw diffraction profiles were consistent with the known binding properties of the ligands and suggested that the PPD approach can be used even prior to a full description using stereochemically restrained Rietveld refinement.

Figure 1

A typical powder sample obtained from precipitation of soluble lysozyme with sodium chloride (scale bar = 50 µm).

Figure 2

High-resolution powder diffraction profile enlarged for 2θ smaller than 4.5° (the full pattern is shown in the inset) of a sample of soluble lysozyme precipitated with sodium chloride. Observed intensities are shown as red crosses and calculated and differences curves are shown as green and magenta lines, respectively. The second and sixth tick marks of the reflection positions (shown in black) correspond to hkl 210 and 201, respectively.

Figure 3

Raw diffraction profiles collected in the 2θ region 4–5° (a) and 13–14° (b). Samples 1–3 correspond to native lysozyme powder, samples 4 and 5 to lysozyme mixed with glucose, 6 and 7 to lysozyme mixed with NAG and 8–10 to lysozyme mixed with NAG₃. An arbitrary number of counts were added to the raw intensities for all profiles to generate these overplots, using values of 0, 1500, 3000, 6000, 7500, 9500, 11 500, 14 500, 16 500, 18 500 for samples 1–10 in (a) and values of −4500, −3500, −1500, 1500, 4000, 2750, 6250, 9000, 9500, 13 000 for samples 1–10 in (b).

Figure 4

Correlation coefficients were calculated between all raw powder diffraction profiles measured from three samples of native lysozyme, two samples of lysozyme mixed with glucose, two samples of lysozyme mixed with NAG and three samples of lysozyme mixed with NAG₃. The numbers to the upper right of the diagonal compare pairwise profiles collected in the 2θ range 4–5° corresponding to 10–8 Å Bragg d-spacing. The numbers to the bottom left of the diagonal compare pairwise profiles collected in the 2θ range 13–14° corresponding to ∼3 Å Bragg d-spacing. The standard Pearson correlation coefficient was used with the 2θ angle assigned as the independent value and the number of counts in the powder diffraction profile assigned as the dependant value.