Translating virus evolution into epidemiology

Intracellular bottlenecking permits no more than three tomato yellow leaf curl virus genomes to initiate replication in a single cell

Viruses are constantly subject to natural selection to enrich beneficial mutations and weed out deleterious ones. However, it remains unresolved as to how the phenotypic gains or losses brought about by these mutations cause the viral genomes carrying the very mutations to become more or less numerous. Previous investigations by us and others suggest that viruses with plus strand (+) RNA genomes may compel such selection by bottlenecking the replicating genome copies in each cell to low single digits. Nevertheless, it is unclear if similarly stringent reproductive bottlenecks also occur in cells invaded by DNA viruses. Here we investigated whether tomato yellow leaf curl virus (TYLCV), a small virus with a single-stranded DNA genome, underwent population bottlenecking in cells of its host plants. We engineered a TYLCV genome to produce two replicons that express green fluorescent protein and mCherry, respectively, in a replication-dependent manner. We found that among the cells entered by both replicons, less than 65% replicated both, whereas at least 35% replicated either of them alone. Further probability computation concluded that replication in an average cell was unlikely to have been initiated with more than three replicon genome copies. Furthermore, sequential inoculations unveiled strong mutual exclusions of these two replicons at the intracellular level. In conclusion, the intracellular population of the small DNA virus TYLCV is actively bottlenecked, and such bottlenecking may be a virus-encoded, evolutionarily conserved trait that assures timely selection of new mutations emerging through error-prone replication.

Persistence of a Frameshifting Deletion in SARS-CoV-2 ORF7a for the Duration of a Major Outbreak

Australia experienced widespread COVID-19 outbreaks from infection with the SARS-CoV-2 Delta variant between June 2021 and February 2022. A 17-nucleotide frameshift-inducing deletion in ORF7a rapidly became represented at the consensus level (Delta-ORF7a^Δ17del) in most Australian outbreak cases. Studies from early in the COVID-19 pandemic suggest that frameshift-inducing deletions in ORF7a do not persist for long in the population; therefore, Delta-ORF7a^Δ17del genomes should have disappeared early in the Australian outbreak. In this study, we conducted a retrospective analysis of global Delta genomes to characterise the dynamics of Delta-ORF7a^Δ17del over time, determined the frequency of all ORF7a deletions worldwide, and compared global trends with those of the Australian Delta outbreak. We downloaded all GISAID clade GK Delta genomes and scanned them for deletions in ORF7a. For each deletion we identified, we characterised its frequency, the number of countries it was found in, and how long it persisted. Of the 4,018,216 Delta genomes identified globally, 134,751 (~3.35%) possessed an ORF7a deletion, and ORF7a^Δ17del was the most common. ORF7a^Δ17del was the sole deletion in 28,014 genomes, of which 27,912 (~99.6%) originated from the Australian outbreak. During the outbreak, ~87% of genomes were Delta-ORF7a^Δ17del, and genomes with this deletion were sampled until the outbreak's end. These data demonstrate that, contrary to suggestions early in the COVID-19 pandemic, genomes with frameshifting deletions in ORF7a can persist over long time periods. We suggest that the proliferation of Delta-ORF7a^Δ17del genomes was likely a chance founder effect. Nonetheless, the frequency of ORF7a deletions in SARS-CoV-2 genomes worldwide suggests they might have some benefit for virus transmission.