Between Objectivity and Whim: Nucleic Acid Structural Biology

Abstract

Here we explore the subjectivity intrinsic to macromolecular crystallography, focusing on the hydration/counter-ion region of nucleic acids. Water molecules, monovalent and divalent cations, and polyamines compete for similar or adjacent sites. Many of these species give identical electron distributions (electron density maps). Such scattering iso-types allow one to construct different models that give similar fits of model to data. Even models with different electron densities can give similar fits of model to data because various parameters compensate. The geometries of the coordinating ligands of many species (magnesium excluded) are similar to each other, and are effectively identical within the limitations of macromolecular crystallography. The observed distances are commonly occupancy-weighted averages. In sum, atom-type assignments and occupancies in the hydration/counter region cannot be unambiguously extracted from conventional X-ray diffraction experiments. We give several examples where incorrect atom-type assignments have been revealed by anomalous scattering experiments. Published structures and the database that contains them do not provide realistic representations of subjectivity. Reconsideration of macromolecular model-building protocols may be in order. Anomalous scattering provides information that can allow one to characterize the hydration/counter-ion region with greater accuracy than was previously possible.