Structure-guided design of novel inhibitors targeting the drug-resistant M2 proton channel from pandemic “swine” influenza

Antiviral drugs are essential in the early response to pandemic influenza, whilst effective vaccines are developed. The 2009 pandemic H1N1 (pH1N1) strain, and other currently circulating Influenza A viruses (IAVs), are almost ubiquitously resistant to the licensed antivirals amantadine and rimantadine, which target the M2 proton channel. Amongst the polymorphisms associated with M2 drug resistance, asparagine at position 31 (N31) is the most prevalent. With resistance to neuraminidase-targeted antivirals also on the rise there exists an urgent need to develop new inhibitors targeting resistant strains. Through the generation of a pH1N1 M2 homology in silico model, we have identified the first non-adamantane compounds targeting M2 that are effective against the N31 pH1N1 strain. Controversy exists over the binding site of current adamantane based antivirals to M2, with both lumenal and peripheral sites described in the literature. The novel compounds identified herein using our in silico model were selected based upon a predicted preference for one or other of these binding sites. As such, these compounds represent useful tools to investigate the potential of targeting both sites in combination, which will help to mitigate potential drug resistance. A novel in vitro assay was established to test the functional preferences of these compounds, with the results largely supporting in silico predictions. Combinations of lumenally and peripherally targeted compounds resulted in synergistic effects upon infectious virus titre, whereas this was not the case for combinations of compounds predicted to bind at the same site; in the case of two lumenally targeted compounds, this resulted in antagonistic effects. These data support the presence of two drug target sites within the M2 proton channel complex. Finally, whilst resistance within M2 was readily selected against amantadine analogue M2WJ332, this was not the case for non-adamantane compounds, further emphasising the potential benefits of investigating novel compound classes.