Data-driven multiscale modeling deciphers MOI-dependent dual antiviral mechanisms of OP7

OP7 defective interfering particles are promising antivirals against influenza A virus (IAV), but their antiviral mechanisms are not fully understood. Here, we developed a data-driven multiscale model of IAV and OP7 coinfection calibrated to in vitro human lung cell data. The model predicted, and experiments confirmed, a previously unrecognized multiplicity of infection (MOI)-dependent switch in the relative contribution of two complementary antiviral mechanisms of OP7. For low IAV MOI, OP7-mediated interferon signaling induces the antiviral effector MxA, restricting nuclear import of IAV genomes, while replication interference subsequently reinforces complete suppression of virus replication. For high MOI coinfection, the interferon response established is too slow to contribute to antiviral activity, and virus inhibition is mediated almost exclusively by replication interference. The model further predicted a therapeutic efficacy of OP7 up to 24 h post infection, which was confirmed in subsequent experiments. Following extension, it also reproduced the experimentally defined prophylactic protection for up to 7 days. Altogether, this experimentally validated coinfection model provides a quantitative framework for understanding and rationally optimizing prospective OP7-based antiviral therapies.