Influenza or the flu, which is caused by infection with an influenza virus, is characterized by symptoms such as fever, cough, headache, muscle and joint pain, severe malaise, a sore throat and a runny nose. Most people recover within one week after onset of symptoms.
However, influenza viruses cause three to five million severe cases of illness and approximately 250 000 to 500 000 fatalities worldwide every year, particularly in children,
elderly and people with underlying malignancies or other infections1. Influenza virus is easily transmitted by both direct contact and through the air. The latter is particularly of
importance in crowded areas, as sneezing or coughing produces small virus-containing droplets which spread easily to nearby persons who breathe in these droplets. In the
respiratory tract, influenza viruses infect airway epithelial cells. After entry in these cells, the viral genetic material, in the form of viral ribonucleoproteins (vRNPs), is released into
the cytoplasm and subsequently enters the nucleus. In the nucleus, transcription and replication occurs. The former leads to production of viral mRNA, which is transported to
the cytosol for translation into viral proteins. Some of these newly produced proteins shuttle back to the nucleus where they initiate the replication of the viral genome. The
newly produced viral RNA molecules form new vRNPs which leave the nucleus and associate with structural viral proteins to form new virus particles. After budding, released
viruses can infect neighboring cells and the cycle starts over again2.
More than 50 years ago Lindenmann et al. discovered a gene which conferred resistance against influenza A virus (FLUAV) infection in mice, and which is now known as Mx13-5. The
term Mx signifies myxovirus-resistance, because mice that can express a functional form of this protein are resistant to infection with influenza A virus and other myxoviruses4,5. Quite
some years later, a human ortholog of murine Mx1 was isolated and named MxA6. Hitherto, several Mx orthologs have been described in practically all vertebrates, and they confer
resistance against a whole array of viruses such as Orthomyxoviridae, Rhabdoviridae, and Bunyaviridae (reviewed by Verhelst et al.7). The exact mechanism by which murine Mx1 exerts viral inhibition is largely unknown. Our lab has contributed to the elucidation of the influenza-specific antiviral mechanism by showing that Mx1 interacts with the
nucleoprotein (NP) and polymerase basic 2 (PB2) protein of influenza A viruses and, that in the presence of mouse Mx1 the interaction between the NP and PB2 molecules is
abolished8. We also found evidence that Mx1 might actively disrupt existing influenza A vRNPs9. Being a type I interferon-inducible protein, Mx1 is thought to primarily exert an
innate antiviral effect by reducing FLUAV early after infection, and preventing viral spread through the airways. It is not known whether Mx1 could also fulfill a role in the antiviral
immune cell compartment after a FLUAV infection. The induction by type I IFNs of an antiviral state in antigen presenting cells and in memory T cells has been reported to
directly affect the immune response against a primary and secondary influenza virus infection, respectively10,11. In addition, human DCs rapidly upregulate MxA and thus, at least
in vitro, become resistant to the virus and can sustain antigen presentation12. Most in vivo studies which examine the FLUAV-induced immune response are performed using mouse
strains which do not possess a functional Mx1 locus13. Based on such models, it has been reported that certain immune cell types, i.e. CD103+ dendritic cells and lung resident
memory CD8+ T cells, are protected against FLUAV infection due to an interferon-induced antiviral state10,11. This was the reason for us to hypothesize that Mx1 could also play a role
in the formation of this antiviral state.
To address this question, we set up an infection model wherein we made use of bone marrow chimeric mice. Since most immune cell types have a hematopoietic origin, bone
marrow transfer from mice with a functional Mx1 locus (B6.A2G Mx1+/+) to mice without a functional Mx1 locus (B6.A2G Mx1-/- ), and vice versa, allowed us to study the possible
function of Mx1 in bone marrow-derived cell types. The bone marrow chimeric mice were infected with a high dose of FLUAV. Multiple parameters were examined (body weight, lung
viral titers, viral mRNA and protein levels), and it was apparent that Mx1 expression in bone marrow-derived cell types was not the main factor determining resistance against
FLUAV infection. The driving force in resistance against FLUAV infection was whether or not Mx1 is expressed in stromal cells. This difference between Mx1-/- and Mx1+/+ recipient mice was also noticeable in the evolution of the levels of eosinophils, monocyte-derived
dendritic cells, and alveolar macrophages in the lung.
Results obtained with the FLUAV infection model were not conclusive to address the hypothesis that Mx1 could play a role in bone marrow-derived cell types after viral
infection. Therefore, we tried a second Orthomyxovirus infection model. Thogoto virus (THOV) is a tick-borne virus which, like FLUAV, belongs to the family of Orthomyxoviruses.
Importantly, small rodents are natural hosts for this virus, and THOV is also sensitive to inhibition by Mx114,15. Bone marrow chimeric mice were infected with a high dose of THOV.
Again, multiple parameters were examined (body weight, liver viral titers, viral protein levels, liver pathology). Like in the FLUAV infection model, the main determinant for
resistance against infection is Mx1 expression in the stromal cells. However, irradiated Mx1- /- recipient mice that had received Mx1+/+ bone marrow cells, displayed reduced morbidity from the THOV infection compared with Mx1-/- recipients which received Mx1-/- bone marrow cells as evidenced by the reduced weight loss and liver pathology which was observed for this group of mice. This observation suggests that Mx1 can play an important role in immune cell types after viral infection, although the importance of this role is largely dependent on the infecting virus.
We also addressed the possible contribution of Mx1 in context of a vectored influenza A NP vaccine antigen. Recently, Altenburg et al. examined whether recombinant modified
vaccinia Ankara (rMVA) vaccines which expressed mutated forms of NP would elicit a stronger antigen-specific immune response than rMVA vaccines expressing the wild type
(WT) form of NP (rMVA NPwt). The introduced mutations were intended to enhance cytosolic retention or degradation of the NP molecules. For this, they either mutated the
nuclear localization signal (NLS) (rMVA-NPmut), deleted the NLS (rMVA-NPΔNLS) or fused ubiquitin to NP (rMVA-UbqNP). In vitro, these mutated NP constructs outperformed rMVANPwt in activating NP-specific T cells. However, immunization of C57BL/6 mice with the mutant rMVA-NP constructs did not result in significantly higher NP-specific CD8+ T cell
responses or protection against influenza A virus challenge than the rMVA-NPwt construct16. We reasoned that this might be because the required threshold of processed NP
antigen for a robust CD8+ T cell response may be readily reached by the WT NP and thus difficult to improve further by NP variant constructs. Therefore, we speculated that mice
which do express a functional Mx1 protein, as opposed to C57BL/6 mice, would be a better suited model to test these different rMVA-NP constructs. Mx1, which has been shown to
interact with NP8, could be the additional restriction factor needed to demonstrate the advantage of these mutated NP constructs.
We vaccinated B6.A2G Mx1-/- and B6.A2G Mx1+/+ mice with the different rMVA-NP constructs. One week after the second immunization the NP-specific CD8+ T cell response
was examined by intracellular cytokine staining (ICS) and enzyme-linked immunospot (ELISPOT) assay using blood and spleen. Both ICS and ELISPOT data showed no significant
differences between the mutated and the WT NP constructs in B6.A2G Mx1+/+ mice.
However, ELISPOT data showed a trend that rMVA-NPmut and rMVA-NPΔNLS constructs elicit a stronger CD8+ T cell response than the rMVA-NPwt construct. To further
substantiate the theory that Mx1 might act as a determinant for the induction of NP-specific cellular responses, additional experiments will have to be performed.
The results obtained with the THOV infection model show clearly that Mx1 plays a role of significance in immune cells upon viral infection. Consequently, we could draw two major
conclusions from this thesis. First, when studying an infection model it is imperative to use a well suited combination of host and virus. Since THOV is a natural pathogen of small
rodents, it is ideal to use in a mouse model. Second, additional to the choice of a suitable host-virus combination, we can also conclude that the tropism of the chosen virus is of great importance. When investigating the antiviral role of a protein in a certain cell type, it is essential that the chosen virus infects this cell type. THOV infects myeloid CD11b+ cells17, and was consequently very well suited for our experiments.