Mathematical model of influenza infection suggests JAK–STAT activity drives severe pathologies in juvenile mice

Children are uniquely susceptible to severe influenza infection, with one million children experiencing severe life-threatening disease each year. However, there is little evidence that an underdeveloped immune system or differences in viral loads are responsible, implicating the host inflammatory response as responsible for increased lung injury in juveniles. Here, we used mechanism-based mathematical modeling, age-specific lung immune data from influenza-infected mice, Bayesian statistics, and rigorous Monte Carlo-based methods to identify immune mechanisms that may be differently regulated in juvenile animals. We hypothesized that the immunological mechanisms between juvenile and adult mice are primarily conserved, and that immune response differences arise due to a minimal set of parameter differences. First, we developed and identified parameter bounds for an ordinary differential equation (ODE) model of the innate immune response to influenza infection which capture the dynamic changes of select parameters. Using publicly available juvenile and adult murine data, we then conducted a computational screen of different age-specific model scenarios and evaluated the scenarios using the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) to select the optimal scenario. These results suggest that the rate of production of JAK-STAT pathway activators, like type I IFNs and IL-6, is age-specific. Preconditioned Monte Carlo (PMC) analysis revealed that JAK-STAT activator production is higher in juveniles than adults. Additional simulations suggest antiviral therapeutics may be more effective in juvenile populations. While not significantly suppressing virus replication, age-specific IFN or IL-6 production may be responsible for increased inflammation, lung injury, and mortality observed in juvenile influenza infection.