Inference from RNA interference: Suggestions to our authors

Abstract

Those of you who read Nature must have been astounded in 2008 about an article stating that “Non-targeted (against nonmammalian genes) and targeted (against Vegfa or Vegfr1) siRNA suppressed choroidal neovascularization via cell-surface toll-like receptor 3 (TLR3), its adaptor TRIF, and induction of interferon-γ and interleukin-12. Non-targeted siRNA suppressed dermal neovascularization in mice as effectively as Vegfa siRNA”. These siRNA effects were entirely sequence-independent and classspecific for double strand RNAs longer than 21 bp. In addition to these TLR-mediated various other types of undesired effects have been ascribed to si/shRNAs and guidelines to avoid the various types of off-target effects have been proposed. This is of particular importance when a study uses si/shRNAs for validation of a limited set of events which are components of the large regulatory networks that underlie complex cellular phenomena like cell fate transitions, growth, or cell death. The International Journal of Cancer (IJC) receives many manuscripts where siRNA or shRNA approaches are used to support essential conclusions. Often only one siRNA is used and frequently the existence of off-target effects is not considered at all. To avoid requesting additional experiments before sending out the manuscripts for review, we recommend for cell assays to (1) use multiple si/shRNAs, ideally targeting different exons, and different ‘scrambled’ RNAs as control, (2) make sure to exclude immune recognition (IFN response), and (3) validate key data either by using a ‘rescue’ approach (e.g., by expressing the 30-truncated cDNA in case the si/shRNA targets the 30UTR, by expressing the targeted transcript with a mutated siRNA recognition site from a corresponding cDNA vector) or by using a different technology like a CRISPR/Cas9 knock out or another type of knock-out. Given that the latter can be rather demanding, we ask authors to provide such validation only for key experiments which support their central conclusion. The Editors of IJC would like to point out that they are all active senior scientists with significant track records who know by their own experience how disappointing it is for authors when a paper is returned without peer review. We do not judge the authors’ records nor the reputation of the institute where the work has been done. For us, scientific quality, novelty, and the impact are the sole criteria for decision. It is for this reason that we draw the attention of our authors to the need of proper controls in RNA interference experiments.