Molecular Recognition of tRNA Species using Solid-State Nanopores

Intracellular RNA molecules adopt a wide variety of 3D structures, many of which have been studied using high-resolution structural techniques such as X-ray crystallography and NMR. However, while canonical structural features for a class of RNA molecules are assumed based on such studies, structural variations within a class are extremely difficult to detect, as only a limited set of structures are available. For example, the number of unique transfer-RNA (tRNA) molecules in eukaryotes can be ∼100, not including variant products of post-transcriptional modification. It has been hypothesized that tRNA accommodation in the ribosome during translation is modulated by tRNA stiffness, which in turn affects the translation kinetics. Despite this, no experimental studies that probe tRNA stiffness are available. In this work, we electrophoretically drive individual tRNA molecules through ∼3nm diameter pores in order to study their deformation kinetics during translocation. Surprisingly, we find that all six tRNAs studied here exhibit unique translocation kinetics, which we believe correspond to variations in their mechanical properties. Implementation of machine learning algorithms reveal a multidimensional set of parameters that can be used to ascertain the identity of a tRNA from a single ionic current pulse with 70-90% accuracy. Our results pave the way for a deeper understanding of RNA tertiary structure and its dependence on post-transcriptional modifications.