by Gertrud U. Rey

Errors during viral replication can give rise to shortened and/or rearranged genomic sequences known as “defective viral genomes” (DVGs). Because DVGs often lack critical elements needed for replication and formation of new viral particles, virions containing DVGs can only complete a replication cycle if they co-infect a cell together with respective full-length (i.e., wild type) viruses. To replicate their genomes, DVGs often hijack missing proteins from the wild type viruses, a phenomenon that can result in suppression of wild type virus replication. There is increasing evidence to suggest that this suppression can be exploited for the development of antiviral agents.

Marco Vignuzzi at the Institut Pasteur has been exploring this idea by investigating the antiviral potential of DVGs produced during infection of cells with various viruses. In an effort to capture DVGs as they were formed in cell culture, Vignuzzi and colleagues infected mammalian and mosquito cells with Chikungunya virus, a mosquito-borne virus that causes symptoms similar to those caused by dengue virus. They then isolated newly emerging virions from the cells and used those particles to infect new cells – a cycle that was carried out 10 times in a technique known as serial passaging. Sequencing and quantification of viral genomes isolated from the last passage revealed that the number of DVGs increased by about 100,000 between the first and last passage. The most prevalent DVGs were sorted into four groups, with DVGs within each group having deletions of similar sizes and at similar genomic locations. Three of the groups included mostly DVGs derived from mammalian cells and the fourth group included mostly DVGs derived from mosquito cells.

To determine whether DVGs are also generated in an infected arthropod, the authors infected Aedes aegypti mosquitoes with Chikungunya virus by allowing them to feed on virus-infected blood. Ten days after infection, the mosquitoes were dissected, total RNA was isolated, and DVGs were identified by sequencing the RNA using DVG-specific primers. This analysis revealed that the DVGs produced in mosquitoes had similar deletion patterns and profiles as those produced in cell culture, suggesting that DVGs are generated both in cell culture and mosquitoes.

All subsequent studies were done using 20 mammalian- and mosquito-derived DVGs that occurred most frequently and persisted through all passages. To confirm that the DVGs were indeed defective and unable to replicate inside a cell in the absence of full-length virus, the authors introduced (i.e., “transfected”) RNA molecules encoding each DVG into mammalian cells and extracted total RNA from the cells at 8, 20, 28, and 44 hours post-transfection. DVG RNA was then quantified by PCR using primers specific for the respective DVGs. The authors observed that in contrast to wild type virus levels observed in control cells, which increased steadily over time, DVG levels decreased across all time points. This finding confirmed that in the absence of wild type virus DVG RNA was not replicated, but degraded over time.

To see whether the 20 DVGs actually interfered with replication of wild type virus, the authors transfected mammalian cells with a 1:1 ratio of an RNA encoding a DVG and an RNA encoding a full-length fluorescently-tagged wild type Chikungunya virus, so they could monitor the presence of the wild type virus by fluorescence microscopy at various timepoints. Wild type Chikungunya viruses transfected together with most DVG-encoding RNAs continued fluorescing strongly at 48 hours post-transfection, suggesting that most DVGs did not inhibit or reduce the replication of wild type virus when transfected at a 1:1 ratio. However, when the DVG/full-length virus ratio was increased to 10:1, fluorescence of full-length Chikungunya viruses decreased by 10 – 1,000-fold in the presence of almost all DVGs, suggesting that a higher ratio of most DVG RNAs increased the likelihood of these genomes to confiscate needed replication elements from wild type viruses and thus interfere with their reproduction. Interestingly, the smallest DVG with the biggest deletions did not seem to interfere with wild type virus replication, probably because this genome was missing too many elements and could not be adequately compensated by the presence of full-length virus. Overall, these results suggested that both mammalian- and mosquito cell-derived DVGs can interfere with wild type virus replication in mammalian cells. Remarkably, when this experiment was repeated in mosquito cells, most of the mosquito cell-derived DVGs that could inhibit wild type virus replication in mammalian cells were unable to do so in mosquito cells. Although the exact reason for this effect is unclear, it is possible that the prevalence of arthropod-borne viruses in mosquitoes has led these viruses to evolve some resistance to the effects of DVGs during viral replication in mosquitoes.

The authors also found that although most DVGs only inhibited Chikungunya virus strains that were closely related to the strain they were derived from, a small number of DVGs also inhibited more distantly related viruses like Sindbis virus, suggesting that DVGs may be capable of inhibiting a broad range of viruses.

In a final set of experiments aimed to evaluate the ability of the DVGs to prevent viral spread within mosquito hosts, the authors injected mosquitoes with DVG-encoding RNAs, and two days later they infected them with a fluorescently-tagged wild type Chikungunya virus. A control group of mosquitoes was only infected with Chikungunya virus but did not receive any DVGs. Five days after infection, the mosquitoes were killed and analyzed for infection and viral spread by looking for the presence of virus in the midgut and the rest of the body, respectively. All mosquitoes had similar levels of virus in the midgut, whether they had received DVGs or not, suggesting that DVGs had no significant impact on infection. However, mosquitoes that had received DVGs had significantly lower levels of virus in the rest of the body compared to control mosquitoes, suggesting that DVGs can reduce replication and spread of virus in mosquitoes.   

Because all experiments involved delivery of DVGs before or concurrently with wild type virus infection, it is unclear whether DVGs would have any therapeutic effect if they were applied after infection. Presently, the most feasible use for DVGs would be as a vector control strategy by engineering and releasing mosquitoes that are unable to transmit virus. However, considering that DVGs have immunostimulatory potential and their presence in humans correlates with milder disease and better outcome after influenza virus, respiratory syncytial virus, hepatitis C virus, and dengue virus infections, it would be interesting to see if they could be applied as direct therapeutics in humans. Using a combination of lab experiments and computational approaches, Vignuzzi and colleagues identified DVGs with optimal interference activity in a follow-up study. Based on these results, the French biotechnology company Meletios Therapeutics is currently developing a new class of antivirals against Zika virus and Chikungunya virus. This is exciting news, because there are currently no effective antiviral treatments for these two viral infections, and I look forward to following up on these new developments in a future post.