Our results provide the first in vivo evidence that MBL deficiency increases susceptibility to IAV infection. Importantly, the increased infection susceptibility can be improved with rhMBL, as administration of rhMBL to MBL null mice reduced viral infection similar to WT levels. This study also has revealed the presence of MBL in the healthy resting lung. We used affinity purification to isolate MBL from BALF. This procedure was clearly more sensitive than ELISA, in which MBL was detected only after infection . In the previous study, MBL in the BALF was measured by ELISA, which further dilutes protein concentrations in addition to the initial dilution from the fluid used to perform lung lavage . We confirmed that ELISA did not detect MBL in un-concentrated BALF that were also used in the previous study . These results suggest that MBL, a serum protein, most likely leaks into the alveolar space and that MBL also participates in innate immune protection against infection in the lung.
This study demonstrates that MBL inhibits viral infection directly as well as indirectly with cooperation of serum factors. These findings support previous studies showing that MBL directly neutralizes and inhibits IAV infection and that there are direct  and indirect anti-viral activities, including involvement of complement . Of interest, we show that Phil82BS, which lacks one glycosylation site from the parent strain Phil82, also activates the lectin complement pathway and thrombin-like activity despite reduced MBL binding. Robust complement activation despite reduced binding of MBL to Phil82BS could be explained by recent findings that the lectin pathway activity is amplified by the alternative pathway, suggesting that even a lower degree of binding may be sufficient in inducing effective anti-viral activity [32, 33].
The sugar specificity of MBL-A and MBL-C is slightly different and MBL-A is an acute phase protein while the expression of MBL-C is not influenced to the same extent by inflammatory stimuli . We now show that MBL-C is more effective in direct anti-viral activity than MBL-A in vitro. This is the first observation demonstrating a difference between MBL-A and MBL-C in inhibiting a pathogen. This difference is diminished by co-operation of serum factors, since MBL-C deficient serum, which is MBL-A sufficient, is as effective as MBL-A deficient serum (MBL-C sufficient) at neutralizing IAV. This cannot be attributed to an increase of MBL-A in MBL-C null mice because our previous study demonstrates that MBL-A in MBL-C null mice is similar to that in WT mice and vice versa . Importantly, these serum-facilitated anti-viral activities are initiated by MBL because MBL null mice serum does not show viral neutralizing activity even at high concentration.
MBL-ligand binding induces conformational changes in MASPs, resulting in activated serine proteases. Surprisingly, thrombin-like activity is mediated by MBL-C whereas the lectin complement pathway is more efficiently mediated by MBL-A (supporting our previous findings ). These data suggest that direct anti-viral activity of MBL-C correlates with thrombin-like activity. Interestingly, human MBL is genetically homologous to MBL-C and also mediates thrombin-like activity . These findings raise the possibility that, in addition to mediating complement activation, MBL may contribute to host defense by activating coagulation. Hence, complement and coagulation activity may be effective innate immune mechanisms not only in primitive animals, like the horseshoe crab,  but also in mammals.
Our study also demonstrates that MBL modulates cellular responses, increasing recruitment of WBCs, and in particular PMNs, which we have shown mediate viral clearance , although the overall predominant cell type is MΦ in both WT and MBL null mice. A marked increase of apoptotic cells was observed in MBL null mice during IAV infection. This result could be explained, in part, by reduced clearance of apoptotic cells, as MBL null mice have impaired apoptotic cell removal . An unexpected finding is that MΦs of MBL null mice seem to be susceptible to apoptosis once these are isolated and placed in vitro.
Pathogenesis of IAV infection has been linked to polymerase basic (PB)1-F2, which induces apoptosis upon infection , suggesting that viral infection induces host cell apoptosis to minimize host cellular responses to the virus. In this scenario, prevention of apoptosis is a host defense mechanism. Taken together, these observations suggest that immune cells of MBL deficient hosts are more easily infected and more prone to apoptosis, and that impaired clearance of apoptotic cells would further increase the burden of infection.
Multifactorial high throughput assays of BALF have revealed that the lung of WT mice is relatively quiescent compared with that of MBL null mice, because only two molecules, IL-3 and eotaxin were increased following IAV infection. Although IL-3, a mast cell growth factor, has been linked to lung diseases in animal studies , mast cells themselves have been known to play a role in wound healing . Eotaxin (CCL11) has been identified in lung tissue repair as a chemo-attractant of airway smooth muscle and as a lung fibroblast growth factor . These observations suggest that the lungs of WT mice are in the wound-healing phase as early as on day 1 after IAV infection. In contrast, the lungs of MBL null mice have increases of 9 molecules: Leptin, Leptin receptor, IFN-γ, IL-1α, PF4, IGFBP-6, P-selectin, VCAM, and Axl tyrosine kinase. Surprisingly, all these molecules have been associated with and/or attributed to lung injuries [39–48], suggesting that MBL deficient hosts may be prone to tissue damage from infection. Moreover, these factors may also contribute to increased susceptibility to apoptosis of MBL null macrophages, as discussed above. Taken together, these observations suggest that MBL plays a role in preventing tissue injury, and further study is required to elucidate the details of these processes.
It is important to note that even MBL deficient mice cleared virus by day 4 in our study. The likelihood is that lung-surfactant proteins are contributing to anti-viral activity, as SP-A and SP-D are synthesized in the lung and possess anti-viral activities, including neutralization, opsonization and hemagglutination-inhibition of virus . Mice lacking SP-A or SP-D are susceptible to IAV infection . Although SP-A, SP-D and MBL belong to the collectin family, the surfactant proteins do not activate complement, contrast to MBL . Surfactant proteins do not seem to form a complex with complement activating serine proteases, such as MASPs, and most likely do not activate coagulation. In contrast to these differences, these three collectins do influence adaptive immunity [49–51] although their influence and the details of their actions against IAV are not well understood. Taken together, these observations suggest that collectins may function cooperatively together to eliminate virus. Further studies are warranted to elucidate the details of the interaction among these collectins.
In conclusion, our study demonstrates in vivo evidence that MBL protects hosts from IAV infection and that MBL may be a new useful adjunctive anti-IAV therapy. Anti-IAV mechanisms include activation of the lectin complement pathway and of coagulation through a thrombin-like activity, both of which are innate immune mechanisms. Our investigation also suggests that MBL deficiency may be a risk factor for IAV infection. Thus, MBL, as an element of the innate immune system, plays an important role in protecting and maintaining lung homeostasis.