The transcriptional and translational landscape of equine torovirus

Hazel Stewart; Katherine Brown; Adam M. Dinan; Nerea Irigoyen; Eric J. Snijder; Andrew E. Firth

This indicates that, contrary to previous reports, low levels of discontinuous RNA 238 synthesis are used during production of the N gene negative-strand RNA. The final 239 chimaera which included the 6 nt leader was represented by three reads. These 240 reads included 44 -46 nt of the 5' UTR (i.e., significantly more than the predicted 241 leader-TRS) followed by a sequence mapping to position 19987-19989 which is 242 within ORF1b. 243 A substantial number of additional chimaeric reads were identified, indicative of 244 non-TRS-driven cases of discontinuous RNA synthesis, although formally it is possible 245 that some of these are template-switching artefacts introduced during library 246 preparation and/or sequencing. Additionally, a large number of reads spanning from 247 the 5' UTR to either within the N protein gene or the 3' UTR were identified. Indeed, 248 the only junction represented by over 1000 reads spanned nucleotides 673 to 27649; 249 similarly the second most commonly identified junction spanned 687 to 27550 (642 250 reads). If chimaeric reads were predominantly a sequencing artefact, the abundance 251 of any particular chimaera would be approximately proportional to the product of 252 . CC-BY 4.0 International license author/funder. It is made available under a The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/296996 doi: bioRxiv preprint the abundances of the sequences from which the 5' and 3' ends of the chimaera are 253 derived (with some variation due to sequence-specific biases), and thus a high 254 density of chimaeras would be expected to fall entirely within the N transcript. In 255 contrast, most of the observed chimaeric reads were between N and the 5' UTR. The 256 relative paucity of reads mapping to generic locations in the ORF1ab region also 257 argues against the majority of chimaeras being simply artefactual. The 5' UTR 258 preference may be due to genome circularisation during negative-sense synthesis as 259 has been proposed for coronaviruses (18). Alternatively these may derive from 260 autonomously replicating defective interfering RNAs, rather than multiple 261 independent RNA translocation and reinitiation events. Such defective interfering 262