The quick and the dead: A guide to fast phasing of small ribozyme and riboswitch crystal structures

Abstract

Ribozymes and riboswitches are examples of non-protein-coding (nc)RNA molecules that achieve biological activity by adopting complex three-dimensional folds. Visualization of such molecules at nearatomic resolution can enhance our understanding of how chemical groups are organized spatially, thereby providing novel insight into function. This approach has its challenges, which mainly entail sample crystallization followed by the application of empirical, structure-determination methods that often include experimental “phasing” of X-ray diffraction data. A paucity of high-quality crystals or a low symmetry space group are factors that demand rapid assessment of phasing potential during an ongoing experiment in order to assure a successful outcome. Here we describe the process of evaluating the anomalous signalto- noise as a prelude to single wavelength or multiwavelength anomalous diffraction (SAD or MAD) phasing. Test cases include an autolytic 62-mer RNA enzyme known as the hairpin ribozyme, and a 33-mer riboswitch that binds the modified guanine metabolite preQ1. The crystals were derivatized with iridium (III) hexammine and osmium (III) pentaammine triflate, respectively. Each data set was then subjected to the XPREP and SHELX programs to assess the anomalous signal-to-noise and to locate the heavy-atom substructure. Subsequent noise filtering was conducted in SHELXE or RESOLVE. The methods described are applicable to the rapid phasing of RNA X-ray diffraction data, and contrast the efficacy of in-house X-rays with those attainable from synchrotron-radiation sources in terms of the potential to plan for and execute an experimental structure determination.