A 39.8kb flavi-like virus uses a novel strategy for overcoming the RNA virus error threshold

It is commonly held that there is a fundamental relationship between genome size and error rate, manifest as a notional "error threshold" that sets an upper limit on genome sizes. The genome sizes of RNA viruses, which have intrinsically high mutation rates due to a lack of mechanisms for error correction, must therefore be small to avoid accumulating an excessive number of deleterious mutations that will ultimately lead to population extinction. The proposed exceptions to this evolutionary rule are RNA viruses from the order Nidovirales (such as coronaviruses) that encode an error correcting exonuclease, enabling them to reach genome lengths greater than 40kb. The recent discovery of large genome flavi-like viruses (Flaviviridae), which comprise genomes up to 27kb in length yet seemingly do not encode exonuclease domains, has led to the proposal that a proofreading mechanism is required to facilitate the expansion of RNA virus genomes above 30kb. Herein, we describe a 39.8kb flavi-like virus identified in a Haliclona sponge metatranscriptome that does not encode an exonuclease. Structural analysis revealed that this virus may have instead captured bacterial domains associated with nucleic acid metabolism that have not been previously found in RNA viruses. Phylogenetic analysis placed this virus as a divergent pesti-like lineage, such that we have provisionally termed it Maximus pesti-like virus. This virus represents the first instance of a flavi-like virus achieving a genome size comparable to that of the Nidovirales and demonstrates that RNA viruses have evolved multiple solutions to overcome the error threshold.