The role of the poly(A) tract in the replication and virulence of tick-borne encephalitis virus

Abstract

The tick-borne encephalitis virus (TBEV) is a flavivirus transmitted to humans, usually via tick bites. The virus causes tick-borne encephalitis (TBE) in humans, and symptoms range from mild flu-like symptoms to severe and long-lasting sequelae, including permanent brain damage. It has been suggested that within the population of viruses transmitted to the mammalian host, quasispecies with neurotropic properties might become dominant in the host resulting in neurological symptoms. We previously demonstrated the existence of TBEV variants with variable poly(A) tracts within a single blood-fed tick. To characterize the role of the poly(A) tract in TBEV replication and virulence, we generated infectious clones of Torö-2003 with the wild-type (A)₃C(A)₆ sequence (Torö-6A) or with a modified (A)₃C(A)₃₈ sequence (Torö-38A). Torö-38A replicated poorly compared to Torö-6A in cell culture, but Torö-38A was more virulent than Torö-6A in a mouse model of TBE. Next-generation sequencing of TBEV genomes after passaging in cell culture and/or mouse brain revealed mutations in specific genomic regions and the presence of quasispecies that might contribute to the observed differences in virulence. These data suggest a role for quasispecies development within the poly(A) tract as a virulence determinant for TBEV in mice.

Figure 1. Characterization of infectious clone in vitro.

(A) Schematic view of the strategy used to rescue Torö-6A and Torö-38A infectious clones. A DNA fragment containing the 5′ NCR, the structural genes, and 124 bp of NS1 was excised from its clone in the pcDNA3.1 vector. Another fragment comprising the non-structural proteins (NS1–5) and the 3′ NCR containing 6 or 38 adenine nucleotides was excised from Torö-2003 luciferase sub genomic replicons. Both fragments were ligated at the pspOMI restriction site, and PCR was used to introduce specific 3′ termini and a SP6 promoter preceding the 5′ NCR. A mixture of HEK293 and Vero B4 cells was transfected with the in vitro-transcribed RNA to rescue the infectious clones. (B) Immunofluorescence assay of Torö-6A or Torö-38A–infected A549 cells using primary antibody against the E protein and Alexa 488 as the secondary antibody. (C) Plaque morphology of Torö-6A or Torö-38A in Vero B4 cells. (D) Vero B4 cells (n = 3) were infected with Torö-6A and Torö-38A at an MOI of 0.1, and viral RNA levels were quantified by real-time RT-PCR at the indicated time points. The viral RNA was standardized to the mRNA levels of GAPDH and presented as the mean ± SEM. Statistical significance calculated by ordinary 2-way ANOVA P < 0.0001. a.u., arbitrary unit.

Figure 2. Pathogenicity of Torö-6A and Torö-38A in vivo.

C57BL/6 mice (n = 10) were inoculated (A) intraperitoneally (10⁴ pfu/mouse) or (B) intracranially (10 pfu/mouse) with the indicated virus. The survival rate was then monitored. Survival differences were tested for statistical significance with the log-rank (Mantel–Cox) test, **p < 0.01.

Figure 3. Integrative Genomics Viewer alignment of next generation sequencing reads of Torö-6A and Torö-38A isolated from mouse brains.

Schematic view of the complete genome of Torö-2003 is illustrated above the alignment. The vertical coloured lines in the coverage plots correspond to nucleotides that differ compared to the reference genome. The alignment is coloured by pair orientation where grey horizontal bars are properly aligned read pairs and green bars are read pairs that differ from the expected orientation. The poor coverage across the poly(A) tract of Torö-38A is indicated by a red arrow.

Figure 4. Alignment of 3′ NCR partial sequences from pcDNA3.1 clones of the rescued viruses.

The sequences correspond to Torö-6A and Torö-38A from cell culture (A,C) and mouse brain (B,D). The number of sequenced clones with identical sequences is shown in parentheses. Nucleotide number 10440 corresponds to the TBEV strain Torö-2003. The poly(A) tract in the reference sequence is highlighted by a green line.

Figure 5. Genomic diversity across the open reading frames of Torö-6A and Torö-38A from cell culture and mice.

For each gene, the number of total mutations was divided by the number of total nucleotides sequenced. Mutation frequencies are presented as mutations per 10,000 nucleotides. Error bars represent the standard error of the mean.

Figure 6. RNA secondary structure prediction of the Torö genomes.

The upper panel illustrates RNA folding of Torö-6A and Torö-38A sequences at 37 °C using mfold (Zuker 2003). The lower panel illustrates folding of the 706 proximal and 701/733 terminal bases to enhance the resolution of the secondary structures in the 3′ and 5′NCRs. (A,B) Present folding of Torö-6A and the variant with a guanine deletion at nucleotide 10495, respectively. Folding of Torö-38A (C), the nucleotides 10527–10592 deletion (D), and the nucleotides 10442–10599 deletion (E) are also depicted. RNA folding predictions show that stem loop (SL) 14 is the only RNA secondary structure affected by the observed deletions and the poly(A) sequence of Torö-38A.