observed for 2 aminobenzenesulfonamide whereas rilmenidine was ineffective. The other molecules had comparable effects on the two H1N1 virus strains, with brinzolamide, midodrine and ribavirin being the most effective Tosedostat Tosedostat Androgen receptor inhibitor Androgen receptor inhibitor antivirals. The EC50 of ribavirin were comprised between 61 mM and 292 mM revealing a resistance to this molecule that was 4 to 10 times more in the H1N1 SOIV strain compared to the H1N1 strain. We compared drug sensitivities to viral growth curves of different viruses after infection of A549 cells at two moi. Viruses with good replication efficiencies and the faster kinetics were the most resistant to the drug panel. In contrast, selected antivirals had a better effect on delayed replication viruses.
Drug sensitivities therefore partially correlated with viral growth kinetics.
However, some strain specificity may also account for drug sensitivities. Indeed, H3N2 virus was one of the most drug sensitive virus, while replicating as efficiently than H7N1 virus. DCC-2036 DCC-2036 To conclude, five molecules out of the eight potential molecules selected by our in silico screening inhibited viral growth of the H1N1 SOIV, a virus that was unknown when we first defined the signature of infection and queried the Connectivity Map. These results are promising and strongly indicate that this approach identifies molecules with a broad anti influenza spectrum of activity.
Discussion The virally induced gene expression signature Influenza infection induces various intracellular signaling cascades and important downstream gene expression host cell modifications.
Despite their host range restriction that may reflect the better adaptation to host factors, all influenza A viruses can infect the same cells in vitro, prompting us to assume that they may hijack common cellular proteins for their own replication. This is the first study to compare the cellular gene expression modifications induced by five different influenza A virus subtypes. As already described in previous transcriptional in vitro and in vivo studies, we found that H5N1 infection induced a strong upregulation of interferon response genes.
This sustained hyperinduction has been correlated with the high virulence of this virus in animal models. In patients, H5N1 infection results in a massive production of cytokines and chemokines, referred to as the cytokine storm, which could be responsible for the severity of the disease.
Here we observed that H5N1 induced the expression of more, and to a greater extent, inflammatory/immune response genes than any of the other subtypes. Molecular mechanisms supporting the higher activation of interferon signaling by H5N1 in comparison with other subtypes remain undetermined. In contrast, we found that A/New Caledonia/20/99 infection leads to the smallest change in gene expression at 24 hpi. One could speculate that H1N1 virus, as a human influenza virus, would be well adapted to human A549 cells and could replicate in these cells with basal level of proteins, thus without having to induce much gene expression changes. However a well adapted virus would efficiently replicate in these cells. We performed replication kinetics in A549 cells with the different viruses and observed that H1N1 virus grew to lower titers than other viruses. Two hypothesis can be formulate