Sequence-Independent RNA Sensing in Living Mammalian Cells

Recently, several groups described sensors in living cells that take advantage of adenosine deaminases acting on RNA (ADARs) to link the presence of an RNA (a "target transcript") to the translation of a payload from a second, exogenously introduced mRNA. These sensors share the key mechanism of editing a stop codon opposite a specific sequence motif in the target transcript, where this motif requirement is dictated by ADAR's strong sequence preference. This constrains sensor design and precludes the sensing of short sequences that lack such motifs, often essential for key applications such as sensing viral RNAs and differentiating splice isoforms. Here we address this limitation with modular RNA sensors using adenosine deaminases acting on RNA ("modulADAR"). ModulADAR features two key elements that mirror the modularity of ADARs: regions that hybridize with the target transcript to recruit ADAR's dsRNA-binding domains, and a stem-loop for stop-codon editing by ADAR's catalytic domain. We optimize modulADAR and apply it to detect short subsequences that cannot be sensed by prior-generation sensors. We anticipate that modulADAR will empower broader basic science and therapeutic applications, especially those that will uniquely benefit from programmable RNA detection in living cells.