Rift Valley fever virus genome organisation: it's like it for a reason

Following recent Bluetongue virus and Schmallenberg virus (SBV) incursions into Central and Northern Europe, Rift Valley Fever virus (RVFV) is now perceived as one of the greatest threats to Europe, crossing over the Mediterranean from Africa where it remains endemic. With good reason. Both Culex  and Aedes species of mosquito are capable of transmitting the virus - West Nile virus infections in Italy have previously shown that there are conditions in Europe suitable for arbovirus transmission by such species of mosquito.
Like SBV, which has spread across Europe in the last 2-3 years, RVFV is a member of the Bunyaviridae family, (albeit in the genus Phlebovirus as opposed to Orthobunyavirus). If it were to encroach into Europe, the headline fact is that it's a zoonosis, with the potential to cause fatal disease in humans. Of equal, if not more, importance is the that fact that it can cause deaths and abortions in ruminants.

Dead cattle as a result of RVFV infection (http://www.elsenburg.com/info/els/077/077e.html)

The RVFV genome comprises three segments of (-)ssRNA. Although the small (S) segment encodes both the nucleocapsid (N) protein and the non-structural protein NSs, a peculiarity is that, whilst N is produced from RNA transcribed from the genome, the message for NSs is transcribed from the antigenome. Superficially this is a little counter-intuitive; NSs is responsible for the rapid abrogation of the interferon system and thus would, presumably, be required as early as possible upon infection. Why force the virus to produce the antigenome before NSs can be produced?

A recent paper by Brennan, Welch and Elliott has addressed this by swapping around N and NSs, such that NSs is translated from the antigenome, and N from the RNA transcribed from the antigenome.

Generation of the swap virus: the (anti)genomic strand from which N and NSs are transcribed is 'swapped'.

In addition to the virus with wild-type (vaccine strain, MP12) and 'swap' virus, they also made viruses where NSs was substituted with EGFP. Now they had a panel of viruses including ones where NSs (or EGFP) were expressed first, before N. When they tested the growth of each virus, the swap viruses all grew poorly compared to the wild-type. This was regardless of whether the cells were either interferon competent (A549), incompetent (BHK21) mammalian cells, or indeed various insect cell lines.

Growth of 'swap' vs wt MP12 RVFV: In all cases, all swap viruses (filled circles) grow with lower efficiency compared to MP12 (A:mammalian cells; C. insect cells).

The lower apparent rate of replication was reflected in the amount of protein. As time progressed, the amount of protein accumulated by the swap virus was lower, although as expected the NSs protein was on this occasion produced before the N protein; opposite to what occurs with the wt virus. The accumulations of NSs were also much more substantial than with wt. However, although there was much more NSs (and less N) than wt, significant amounts of the glycoprotein (Gn) were not detected until 48 hours after infection, compared to 18-24 in the case of wt. In the case of the swap virus with EGFP in place of NSs, Gc was virtually undetectable even at 48h.

In wt RVFV, NSs assembles into filaments that are localised to the nucleus. Rather oddly, in the case of the swap virus, these were much thicker than the wt, with additional evidence of some NSs in the cytoplasm. These alterations in NSs behaviour appeared not to affect the functions of NSs in either inhibiting both host protein synthesis and host RNA synthesis. Given its 'shut-off' function, it is intriguing that increasing the abundance of NSs had no additional effect upon the intensity or rate at which protein shut-off occurs.  

Shut-off of host cell macromolecular synthesis: A: radioactive methionine/cysteine incorporation at different times post-infection; less label =  less protein = shut-off. B: shut-off of RNA synthesis (newly synthesised RNA is green, red is virus).

When they looked at the targets of NSs that result in shut-off, p62 was inhibitted by the swap virus (although more slowly), just as the wt but, bizarrely, PKR appeared to be largely unaffected by the swap virus, in contrast to the wt virus where PKR levels dropped from around 5 hours onwards. Overall, it would seem that, whilst it still does, the swap virus is a bit less efficient at the shut-off. As a result, it is a little surprising that the swap virus results in the induction of less IFN than the wt virus in A549 cells.

One thing the authors did tease apart is the relative number of genome:antigenome copies in both the cell and virion fractions. The swap virus was found to transcribe much more NSs RNA compared to the wt equivalent, suggesting that increased activity of the relative promoters is responsible for the dramatic amount of NSs observed with the swap virus. Such an excess of RNA may have overwhelmed the control by RNAi in insect cells, resulting in cytopathogenic effect in infected insect cells (in contrast to wt virus, which establishes a persistent infection). One interesting finding is that more antigenome than genome copies are packaged in swap virus virions. Does this reflect the abrogation a specific packaging process, or simply the abundance of genome:antigenome copies in the cytoplasm when packaging occurs? Considering the swap virus has N transcribed from the antigenome, the increase in antigenome packaging may actually overcome some of the temporal regulation achieved by swapping the N from the genomic to the antigenomic transcipt. It seems a bit more work  is required here. Virions have been shown to contain just 3 segments of RNA. If more of the virions have an antigenomic S segment, then the number of particles per infectious unit will necessarily be increased. It may be that such differences in the make-up of the viral population can explain some of the characteristics of the swap virus.

Of course such attenuating features - at least in vitro - may contribute towards the rational development of a vaccine. The most tempting feature of the swap virus described here is the fact that the virus cannot persist in insect cells and would thus be somewhat resistance to transmission; a key consideration in live arbovirus vaccines.

So, genome organisation is important. Ultimately this is not surprising: there's a reason it's like it is.

Brennan, B., Welch, S., & Elliott, R. (2014). The Consequences of Reconfiguring the Ambisense S Genome Segment of Rift Valley Fever Virus on Viral Replication in Mammalian and Mosquito Cells and for Genome Packaging PLoS Pathogens, 10 (2) DOI: 10.1371/journal.ppat.1003922