Recovery from influenza virus infection has long been known to require an intact T-cell compartment. More recent studies revealed that CD8 and CD4 T cells can promote recovery through independent mechanisms. The CD4 T-cell-dependent recovery process appears to operate primarily through promotion of the T-dependent antibody response as B-cell-deficient microMT mice cannot recover from infection if they have been depleted of CD8 T cells. The potential therapeutic activity of the B-cell response was further studied by transfer of antibodies into infected SCID mice. At the dose of 200 micrograms/mouse, most antibodies (of IgG2a isotype) to the viral transmembrane protein HA cured the infection, while those to the transmembrane proteins NA and M2 suppressed virus titers in the lung but failed to clear the infection. The ability of passive antibody to resolve the infection was closely related to its prophylactic activity, suggesting that neutralization of progeny virus (VN) played an important role in the process of virus clearance in vivo, while reaction of antibodies with infected host cells contributed to but was insufficient, on its own, for cure. HA-specific antibodies of IgM and IgA isotypes were therapeutically ineffective against pulmonary infection, presumably because of a preferential delivery into the upper respiratory tract, while IgG exhibited highest activity against pulmonary and minimal activity against nasal infection. B cells appear to be of similar importance for recovery from primary infection as CD8 T cells.

No abstract

Seasonal epidemics of influenza virus result in ∼36,000 deaths annually in the United States. Current vaccines against influenza virus elicit an antibody response specific for the envelope glycoproteins. However, high mutation rates result in the emergence of new viral serotypes, which elude neutralization by preexisting antibodies. T lymphocytes have been reported to be capable of mediating heterosubtypic protection through recognition of internal, more conserved, influenza virus proteins. Here, we demonstrate using a recombinant influenza virus expressing the LCMV GP33-41 epitope that influenza virus-specific CD8+ T cells and virus-specific non-neutralizing antibodies each are relatively ineffective at conferring heterosubtypic protective immunity alone. However, when combined virus-specific CD8 T cells and non-neutralizing antibodies cooperatively elicit robust protective immunity. This synergistic improvement in protective immunity is dependent, at least in part, on alveolar macrophages and/or other lung phagocytes. Overall, our studies suggest that an influenza vaccine capable of eliciting both CD8+ T cells and antibodies specific for highly conserved influenza proteins may be able to provide heterosubtypic protection in humans, and act as the basis for a potential “universal” vaccine.

Antibodies (Abs) can contribute to the cure of a viral infection, in principle, in two ways by: (1) binding to infected cells and thereby reducing the production of progeny virus [here termed cell-targeting (CT) activity] and (2) reacting with released progeny virus and thereby inhibiting the spread of the infection [termed virus neutralizing (VN) activity]. We have previously shown that a pulmonary influenza virus infection in severe combined immunodeficient mice could be cured by treatment of these mice with hemagglutinin (HA)-specific monoclonal Abs (mAbs) that mediated both of the above activities. Although the therapeutic activity of these mAbs correlated with their VN activity, it remained unclear how much their CT activity contributed to the Ab-mediated recovery process. To clarify this point, we tested the therapeutic efficacy of two mAbs of IgG2a isotype that mediated CT but no VN activity: one specific for the viral neuraminidase and the other for matrix protein 2. Both mAbs reduced pulmonary virus titers by 100- to 1000-fold but they failed to clear the infection, even when administered in combination and at therapeutically saturating concentrations. The results suggest that CT activity contributes significantly also to the therapeutic activity of HA-specific mAbs and further support the notion that VN-activity is required for Ab-mediated virus clearance.

The ectodomain of matrix protein 2 (M2e) of human influenza type A virus strains has remained remarkably conserved since 1918. Because M2e-specific immunity has been shown to decrease morbidity and mortality associated with influenza virus infection in several animal models and because natural infection and current vaccines do not appear to induce a good M2e-specific antibody (Ab) response, M2e has been considered as potential vaccine for inducing cross-reactive protection against influenza type A viruses. The high degree of structural conservation of M2e could in part be the consequence of a poor M2e-specific Ab response and thus the absence of pressure for change. To assess this possibility, we studied the course of infection in SCID mice in the presence or absence of passive M2e-specific monoclonal Abs (MAbs). We found that virus mutants with antigenic changes in M2e emerged in 65% of virus-infected mice treated with M2e-specific but not control MAbs. However, the diversity of escape mutants was highly restricted since only two types were isolated from 22 mice, one with a proline-to-leucine and the other with a proline-to-histidine interchange at amino acid position 10 of M2e. The implications of these findings for the use of M2e as a broadly protective vaccine are discussed.Current influenza virus vaccines aim to induce strong antibody (Ab) responses to the ectodomains of hemagglutinin (HA) and neuraminidase (NA) molecules, since these antibodies (Abs) can provide potent protection against infection and/or disease. The main deficiency of this protection is that it targets highly variable viral determinants. This necessitates not only frequent updating of the vaccine to contemporary circulating virus strains but, given that vaccines have to be produced and applied ahead of exposure to epidemic strains, also a correct prediction of these future epidemic strains. Failure to anticipate the emergence of an epidemic strain with significant antigenic changes compared to the vaccine strain will greatly reduce vaccine-induced protection. It would be advantageous, therefore, to expand vaccine-mediated protection to less variable viral targets. One possible way to achieve this may be through induction of 9,10,14,18,21,24,28).M2 is a 97-amino-acid transmembrane protein of influenza type A virus (15, 16). The mature protein forms homotetramers (12, 29) that have pH-inducible ion channel activity (27,29). M2-tetramers are expressed at high density in the plasma membrane of infected cells but are relatively excluded from sites of virus maturation and therefore incorporated only at low frequency into the membrane of mature virus particles (30,33). Most important in the present context are, first, that the sequence of the 24-amino-acid ectodomain of M2 (M2e) has remained remarkably conserved among human epidemic virus strains (Fig. 1A) (20). Indeed, the majority of human epidemic strains isolated since 1918 share the same M2e protein sequence. Second, several studies in mice have shown that M2e-specific Abs restrict influenz...