Functional characterization of 5′ untranslated region (UTR) secondary RNA structures in the replication of tick-borne encephalitis virus in mammalian cells

Abstract

Tick-borne Encephalitis Virus (TBEV) is an emerging flavivirus that causes neurological disorders including viral encephalitis of varying severity. Whilst secondary RNA structures within the 5′ untranslated regions (UTRs) of many flaviviruses determine both virus replication and pathogenic outcomes in humans, these elements have not been systematically investigated for TBEV. In this study, we investigated the role of predicted RNA secondary elements of the first 107 nucleotides (nts) of the viral genome forming the stem-loop A (SLA). Experiments were performed in replicons and infectious TBEV system. This region comprises three distinct structures: 5’ stem 0 (S0), stem-loop 1 (SL1) and stem-loop 2 (SL2). S0 was found to be essential for virus infection as mutations in the lower stem of this region significantly reduced virus replication. Point mutations in SL1 that preserved the Y-shape confirmation delayed viral RNA replication but did not abolish virus infectivity. Deletion of SL2 did not abolish infectivity but had a negligible effect on virus propagation. No correlation was observed between in vitro translation efficiency and virus infectivity, suggesting that the 5’UTR functions independently to virus translation. Together, these findings reveal distinct RNA elements within the 5′UTR that are essential for the stability and replication of viral RNA. We further identify changes in RNA folding that lead to altered TBEV infectivity and pathogenesis.

Untranslated Regions of a Segmented Kindia Tick Virus Genome Are Highly Conserved and Contain Multiple Regulatory Elements for Viral Replication

“…As reported, JMTV retained the 5′-CAAGUG-3′ sequence in the UTR of all the segments during its first isolation, which indicated that all four segments belonged to the same virus [ 2 ]. The conserved location of the 5′-CAAGUG-3′ sequence in the first structures of the untranslated regions may indicate that this region is involved in interactions with viral and/or cellular proteins to regulate the virus’s life cycle, similarly to how it occurs in other viruses of the Orthoflavivirus genus [ 34 , 35 , 36 , 37 , 38 ].…”

The myriad roles of RNA structure in the flavivirus life cycle

“…While the interaction between SLA and NS5 remains one of the best characterized interactions between the flavivirus RNA genome and a viral protein, recent advances have challenged the canonical view of the SLA structure. Traditionally, SLA was thought to take on a ‘Y’ shaped conformation, characterized by a base stem, side stem-loop (SSL) and a top stem-loop (TSL) ( Figures 1 and 2 ) [ 14 , 58 , 59 ]. However, recent crystallographic and cryo-EM-based structural studies suggest that SLA may at least sample an alternative ‘L’ or ‘V’ shaped structure upon NS5 binding ( Figure 2 ) [ 14 , 15 , 60 ].…”

“…These new models for SLA and NS5 binding are largely consistent with known requirements for this interaction. Beyond the overall structural architecture, foot-printing and mutational studies have consistently demonstrated that NS5 binding to SLA also depends on a conserved ‘AG’ motif in the unpaired loop of the TSL that is maintained in all conformations [ 14 , 27 , 51 , 58 , 62 ]. Despite the different angle of the TSL relative to the base of SLA in the ‘L’ and ‘V’ conformations, the AG motif in the unpaired loop of the TSL sits on the same face of the structure, implying that they both interact with their respective NS5 proteins along the same surface ( Figures 2b-c & 2e-f ) [ 14 ].…”

“…However, the significance of this potential interaction in flaviviruses remains unclear ( see Box 3 ) [ 60 ]. Interestingly, while the predicted SSL is relatively short (2–3 bp) in the mosquito-borne flaviviruses, it is predicted to be significantly longer (~20 bp) in the tick-borne flaviviruses [ 58 ]. It is not yet clear what implication this would have on the ability of SLA to arrange into the putative ‘L/V’ conformation and interact with NS5.…”

“…However, it is not clear from these experiments whether this effect is due to genome dimerization or simply the importance of these nucleotides in interactions with NS5 [ 27 ]. Moreover, it remains unclear if this kissing loop interaction can occur in the tick-borne flaviviruses, which have a much longer SSL [ 58 ]. Regardless, it seems clear that one or a few viral genomes is sufficient to establish an infection, suggesting that genome dimer/multimerization is unlikely to be an absolute requirement at least at the early stages of the viral life cycle.…”