Myxomatosis, a lethal disease of rabbits, is caused by the myxoma virus — a member of the poxvirus family. The myxoma virus does not cause disease in any other vertebrate species, but the molecular basis of this strict species restriction — a characteristic of all poxviruses — is still unclear. Now, reporting in Nature Immunology, Grant McFadden and colleagues show that the myxoma virus species barrier is mediated by a host-cell signalling cascade — the extracellular signal-regulated kinase (Erk)1/2–interferon–STAT1 signalling pathway.

Although myxomatosis only affects rabbits, the myxoma virus can infect mouse cells in vitro, providing scientists with model systems to study poxvirus host restriction. So, McFadden and co-workers infected primary mouse embryonic fibroblasts (pMEFs) with myxoma virus that expressed β-galactosidase under the control of a late viral gene promoter. X-gal assays of infected cells showed that the virus had successfully entered the pMEFs, but that the infection had aborted coincident with late viral gene expression — a stage at which the virus interacts with cell signalling pathways.

So, it seems that host-cell signalling cascades might restrict myxoma virus replication and, sure enough, the infection of pMEFs with myxoma virus in the presence of U0126 — an inhibitor of the signalling molecule Erk1/2 — resulted in permissive viral replication. Activated Erk1/2 usually translocates to the nucleus, where it phosphorylates Elk1. Surprisingly, in myxoma-virus-infected pMEFs, Elk1 remained unphosphorylated, and phosphorylated Erk1/2 was retained in the cytoplasm. The authors reasoned that, in myxoma-virus-infected cells, Erk1/2 might mediate its potent antiviral effect by interacting with other cytoplasmic signalling molecules such as interferon regulatory factor 3 (IRF3) — an important component of the type 1 interferon pathway.

A diverse set of experiments showed that this was indeed the case. Myxoma-virus infection of pMEFs was associated with increased levels of type 1 interferon mRNA, which were substantially reduced on inhibition of Erk1/2. Also, the phosphorylation of STAT1, a downstream mediator of interferon 1, was dependent on Erk1/2 activity, and an association between Erk1/2 and IRF3 was revealed — IRF3 was not phosphorylated and was retained in the cytoplasm of myxoma-virus-infected cells on inhibition of Erk1/2.

These in vitro results were complemented by an in vivo demonstration of cross-species infection. Stat1−/− mice were infected with intracranial myxoma and, in contrast to wild-type mice, which are resistant to this rabbit-specific virus, the Stat1−/− mice rapidly succumbed to infection.

As many viruses exploit the Erk1/2 pathway to promote their replication, these studies reveal the complexities of virus–host interactions and show that modulation of the same host-cell signalling pathway by different viruses can have profound effects on the outcome of infection. And immunologists will also be taking note as these studies have shown a “previously unrecognized link between Erk1/2 signalling and type 1 interferon induction”.