Revealing Sources of Variation for Reproducible Imaging of Protein Assemblies by Electron Microscopy.

Electron microscopy plays an important role in the analysis of functional nano-to-microstructures. Substrates and staining procedures present common sources of variation for the analysis. However, systematic investigations on the impact of these sources on data interpretation are lacking. Here we pinpoint key determinants associated with reproducibility issues in the imaging of archetypal protein assemblies, protein shells, and filaments. The effect of staining on the morphological characteristics of the assemblies was assessed to reveal differential features for anisotropic (filaments) and isotropic (shells) forms. Commercial substrates and coatings under the same staining conditions gave comparable results for the same model assembly, while highlighting intrinsic sample variations including the density and heterogenous distribution of assemblies on the substrate surface. With no aberrant or disrupted structures observed, and putative artefacts limited to substrate-associated markings, the study emphasizes that reproducible imaging must correlate with an optimal combination of substrate stability, stain homogeneity, accelerating voltage, and magnification.