Abstract
Ribosome profiling provides genome-wide snapshots of translation dynamics by determining ribosomal positions at sub-codon resolution. To maintain this positional information, the translation inhibitor cycloheximide (CHX) has been widely used to arrest translating ribosomes prior to library preparation. Several studies have reported CHX-induced biases in yeast data casting uncertainty about its continued use and questioning the accuracy of many ribosome profiling studies. However, the presence of these biases has not been investigated comprehensively in organisms other than Saccharomyces cerevisiae. Here, we use a highly standardized and optimized protocol to compare different CHX-treatment conditions in yeast and human cells. Our data suggest that unlike in S. cerevisiae, translating ribosomes in human cells are not susceptible to conformational restrictions by CHX. Furthermore, CHX-induced codon-specific effects on ribosome occupancy are not detectable in human cells nor in other model organisms including Schizosaccharomyces pombe and Candida albicans. In fact, we find that even in S. cerevisiae most biases can be avoided by omitting CHX pre-treatment, indicating that other parameters of library generation contribute to differences between ribosome profiling experiments. Together our findings provide a framework for researchers who plan their own ribosome profiling experiments or who analyze published datasets to draw judicious conclusions.