Rational Design of Novel xrRNAs

Introduction

Structured elements in RNA viruses are key regulators of the viral life cycle that control processes such as transcription, translation, packaging, or genome cyclization. They are found in both coding and non-coding regions and their occurence in viral untranslated regions (UTRs) and have been extensively studied. In many cases, structured RNAs are evolutionarily conserved throughout virus families and genera.

Exoribonuclease-resistant RNAs (xrRNAs) represent an emerging class of structured viral RNAs that can be found at least once in the beginning of the 3’UTR in all flaviviruses. As the name implies, this relatively short motif of about 80 to 100 bases was evolutionarily optimized to halt the progression of exoribonuclease 1 (XRN1). Structural and biophysical studies have shown that this property is encoded by a ring-like fold centered around a three-way junction (3WJ) stabilized by two pseudoknots (PK1 and PK2) and base triplet. This results in a specialized three-dimensional structure, where the 5’-end is threaded through the RNA core, yielding a topologically closed shape that physically blocks exoribonuclease progression. xrRNAs thus represent a natural example of mechanically active RNA nanostructures in which tertiary topology, as opposed to sequence composition, defines biological function. In this project, we employ xrRNA as a benchmark system for rational RNA design.

People

This Project is a collaboration bewteen the TBI at the University of Vienna and the Institute for Biochemistry at Leipzig University.

Supervisors: Dr. Michael T. Wolfinger (Vienna), Prof. Mario Mörl (Leipzig)

PhD students: Leonhard Sidl (Vienna), Jule Walter (Leipzig)

Current Project State

After thorough analysis of sequence and structure of biological xrRNAs, we created a design-model which describes the sequence space containing natural xrRNA sequences. This not only includes sequence conservation and common secondary structure elements, but also correlations between structural motifs and tertiary structure information such as pseudoknots and base-triples. The RNA design process utilized this model as a design blueprint, using the Infrared-framework (Yao et al., 2024) to sample sequences falling into the constraints defined by it. Sampling was followed by Monte Carlo optimization toward the target secondary structure. Subsequently, designs were validated in silico by simulating a nanopore-sensing experiment using molecular dynamics simulations. To this end, the 5’ end of the structure, the attack site of exoribonucleases, was placed above a pore representing the active site of the enzyme, and a force gradient was applied to the 5’ most nucleotide, pulling it through the pore. Using these simulations, the force required to break up the ring structure can be compared between candidates.

We demonstrated that such a rational design strategy for mechanically active RNAs is feasible by engineering synthetic xrRNAs that match the performance of their natural counterparts. While the first designs still retained significant sequence conservation compared to biological xrRNAs, those constraints could be removed for the last design, resulting in a designed sequence with no detectable similarity to known xrRNA. This shows that mechanical performance can be decoupled from biological sequence ancestry and that force resistance is accessible through de novo design based solely on structural and energetic constraints. (Walter et al., 2026)