6. Summary and Future Directions
6.4. Superinfection exclusion may pose a barrier to reovirus reassortment
show that reovirus +RNA is not precluded from accessing existing VFs. While this only represents one of many viruses with segmented genomes, this finding may still have broader implications. For instance, many viruses of the Reoviridae family replicate within proteinaceous VFs that are similar to those established by reovirus (Tenorio et al., 2019). As such, the ability of +RNA from coinfecting viruses to readily access VFs from both viruses may translate to other viruses of the Reoviridae family and may help explain why reassortment is common in the evolutionary history of many of these viruses. More broadly, this may highlight the importance of viruses to be able recruit RNA to sites of replication for reassortment to occur. Thus, future investigations into RNA localization from coinfecting viruses with segmented genomes that utilize distinct replication strategies from reovirus could reveal whether all viruses with segmented genomes allow RNA from coinfecting viruses to co-localize with replication
compartments.
capable of reassorting genome segments during superinfection, and I show that this occurs even in the absence of segment mismatch (Fig. 3-3C). However, I also show that superinfection is suppressed following primary infection with a type 3, and to a lesser extent type 1, reovirus (Fig. 3-7B-C; Fig. 4-1). These studies were the first to exhibit that reovirus suppresses superinfection by closely related viruses and may have important implications for reovirus reassortment.
In Chapter 4, I begin to assess the mechanism of reovirus-mediated superinfection exclusion. Given that reovirus induces the expression of type 1 interferon, this seemed an obvious candidate for potentially driving superinfection exclusion. My in vitro studies indicated that suppression of secondary virus replication is minimal in type I IFN
receptor knockout cells, suggesting that type I IFN is a major driver of reovirus mediated superinfection exclusion in cell culture. However, in mouse models, type I IFN and type III IFN signaling had no influence on reassortment, regardless of whether coinfection was simultaneous or asynchronous (Fig. 4-2; Fig. 4-3). In fact, reassortant viruses were not detected following a 24-hour time delay to superinfection in mice, even in the
absence of segment mismatch. Similar observations have been made for influenza A virus (Marshall et al., 2013). The observation that reassortants were not detected during superinfection in vivo is not necessarily indicative of superinfection exclusion. Rather, it likely stresses the requirement of coinfection for reassortment to occur, and future studies should directly address whether the absence of reassortment is due to
superinfection exclusion or is a byproduct of how reovirus spreads within a host. This could be determined by quantifying the number of cells infected by a superinfecting virus following a mock primary infection or primary infection with a type 3 reovirus.
Branched DNA FISH probes specific to coinfecting viruses, which were developed as part of this dissertation, paired with flow cytometry would be useful tools in future studies of in vivo coinfection frequency and superinfection exclusion. Specifically, the number of cells that are infected by a superinfecting virus following mock primary infection or 24-hour primary infection of wild-type, type 1 IFN knockout, and type 3 IFN knockout mice with T3DI reovirus could be quantified with branched DNA FISH probes specific to the superinfecting virus by flow cytometry. If superinfection is reduced in wild- type mice and knockout of IFN receptors does not recover the reduction of
superinfection, this would indicate that superinfection exclusion is not mediated by IFN in mice. To further assess the mechanism of superinfection exclusion, antibodies that specifically bind the reovirus receptor JAM-A, or staining for caspase expression, would reveal whether cells in the infected host show reduced expression of the entry receptor, or are undergoing cell death, in response to infection. This may suggest that one of these factors contributes to reovirus superinfection exclusion.
Reassortment frequency can differ depending on the host species (Postnikova et al., 2021; Lin et al., 2017). For instance, influenza A virus frequently reassorts genome segments in avian hosts, but rarely reassorts in humans (Leonard et al., 2017).
Although the precise frequency of reovirus infection and coinfection has not been determined, over half of children have been exposed to type 3 reovirus by the age of 5 (Tai et al., 2005). The finding that reassortment is frequent during coinfection of mice, but was not detected during superinfection, could have important implications for natural infection. If this finding translates to other host species, it may suggest that hosts not only need to be coinfected, but must be coinfected within a narrow time frame, to allow
for robust reassortment. This observation may be indicative of an additional layer of regulation that is required for reassortment to occur during natural infection and may highlight the importance of frequent coinfection as a requirement for reassortment.
Furthermore, within a given host, viruses undergo frequent genetic bottlenecks in infected cells (Zwart and Elena, 2015). Viruses with RNA genomes often have error prone polymerases that lead to the rapid accumulation of mutations, such that viral populations can be thought of as “mutant clouds” made up of a large number of variant genomes (Domingo and Perales, 2019). I have demonstrated that reassortment is most frequent when reoviruses enter cells at the same time and are equally abundant within a coinfected cell. Taken together, this may suggest that within an infected host,
reassortment occurs frequently among highly-similar variants. Whether this is true, and whether reassortment among minor variants confers an advantage to viruses with segmented genomes would be an interesting course of future investigation.