Protection of mice from lethal influenza: evidence that defective interfering virus modulates the immune response and...

Intranasal administration of defective interfering (DI) influenza virus (A/WSN) ensured the survival of 80% of C3H mice otherwise lethally infected with WSN by the intranasal route, whereas a control group which received beta-propiolactone-inactivated DI virus in place of DI virus died at 7.4 days post-infection. DI virus-treated mice developed significantly less lung consolidation than controls although qualitatively the cellular pathology in the two groups was indistinguishable. Surprisingly, in view of the accepted mode of action of DI virus interference, multiplication of infectious virus in the lung, production of viral haemagglutinin (HA) antigen and neuraminidase, and the distribution and amount of viral antigen in cells as shown by immune labelling were unaffected by the presence of active DI virus. Furthermore, assays of lung extracts showed that DI virus was not stimulating significantly greater amounts of interferon than the control inactivated DI virus. An alternative explanation arises from the fact that the pathology of influenza in inbred mice is immune (T lymphocyte)-mediated. Thus, since there is no evidence that DI virus affects virus multiplication we suggest that DI virus is responsible for ameliorating the damaging host responses. Another aspect of the immune response modulated by DI virus was the enhancement of local haemagglutination-inhibiting (HI) antibody in the lung, with peak increases of up to 10-fold over the relevant controls being demonstrated at 5 days after infection. This antibody was presumably complexed to HA antigen in the lung as activity was only demonstrated after elution at low pH. It had no detectable neutralizing activity (less than 10% HI: neutralization ratio of convalescent serum) which accounts for the coexistence of local antibody and virus infectivity. Mice infected with virus alone or which received beta-propiolactone-inactivated DI virus in addition to a lethal dose of WSN did not develop significant amounts of lung antibody. No differences were seen in serum HI titres. The increased level of antibody could not be attributed to the presence of greater amounts of HA antigen in lungs of mice treated with DI virus, as ELISA showed no significant difference from control preparations. The possibility that the two modulated immune responses are linked through HI antibody blocking access of T cells to cell membrane-borne HA antigen is discussed.