Bioinformatics identification of copyback and multihost-adapted defective viral genomes in dengue virus

Defective viral genomes (DVGs) have been detected in clinical samples, and their antiviral effects have been verified. However, due to the limitations of traditional methods, all the reported DENV DVGs to date are Deletion DVGs, and it has not been proven that these DVGs can be stabilized across different hosts.

We used the bioinformatics software DVGfinder with various dengue virus NGS data, including samples from patients, Aedes albopictus, C6/36 cells, and Vero cells, to identify DVGs. We compared the distribution of DVGs across different hosts and analyzed the dynamics of DENV-1 DVGs during serial passaging, identifying DVGs that are stably maintained across hosts.

First of all, we found copyback type DVGs in different datasets, with the dominant DVGs in patient sera being 3' copyback. Secondly, we observed that in C6/36 cells and Aedes albopictus, the DVGs did not change significantly with DENV-1 passages, while in Vero cells, the number of specific Deletion DVGs continuously increased with DENV-1 passage. Finally, using clustering algorithms, we identified a set of candidate deletion DVGs predicted to stably exist across different hosts. One of these candidates, designated DeletionA6 (BP 1320, RI 7700), was experimentally validated by nested PCR in sera from patients infected with DENV-1 to DENV-4.

This study describes the distribution patterns of DVGs across different samples and provides preliminary bioinformatic evidence for a subset of deletion DVGs.