Detection and Quantification of Influenza Virus Defective Viral Genomes from NGS Datasets Obtained after RT or RT-PCR Product Sequencing ^†

Abstract

Like most RNA viruses, influenza viruses (IAV) generate defective viral genomes (DVGs) during viral replication. Although there is accumulating evidence of a biological impact of DVGs, the molecular mechanisms leading to their production remain to be unveiled. Various next-generation sequencing (NGS) technologies and detection methods can be used to characterize DVGs. Here, we developed a bioinformatics pipeline called DG-seq to quickly identify and quantify DVGs in influenza viral stocks and compared two processing methods for NGS, with or without PCR amplification. To evaluate the performance of the DG-seq pipeline, we used either synthetic in-vitro-transcribed DVGs mixed with the full set of synthetic full-length genomic RNAs, or biological RNA samples extracted in duplicate from three IAV stocks: mutant viruses with a K635A or a R638A mutation in the PA subunit of the polymerase that impairs viral transcription, and their wild-type (WT) counterpart. Viral genomic RNAs were reverse-transcribed and either directly subjected to Illumina sequencing (RT-seq) or PCR-amplified prior to sequencing (RT-PCR-seq). Both methods displayed a good reproducibility between batches, with a lower detection rate but a more accurate quantification of DVGs in RT-seq samples. The PA mutants produced more DVGs than the WT virus, derived mostly from the polymerase gene segments, but also from the NA and HA segments, suggesting that an imbalance between transcription and replication can promote DVG production. Breakpoints occurred near the segment extremities, with no hotspot identified. Interestingly, we observed short direct A/T-rich repeats adjacent to the breakpoint ends at a significantly higher frequency than in the random case. This work provides the first comparison of DVG detection and quantification from NGS data obtained in the presence or absence of PCR amplification and gives novel insight into the mechanisms of influenza virus DVG production.