Associate Professor University of Pittsburgh, United States
Introduction: In humans, reproductive-age females are at greater risk than their male, age-matched counterparts for hospitalization and death from influenza infection. During the 2009 H1N1 pandemic, females in the United States made up 53.2% of hospitalizations, versus 46.8% for males, and were at higher risk of death than their male counterparts. The innate immune response has been implicated as a factor of these sex differences in influenza pathogenesis. Females have greater type I interferon activity than males, which are responsible for the induction of IFN-stimulated genes and cause antiviral responses to be activated in infected and neighboring cells during influenza infection. Females have also been shown to have increased production of CCL2/MCP1 during infection, which is a key regulator of monocyte infiltration at the site of infection. Mathematical modeling is a powerful tool for integrating dynamic, sex-specific immune data into testable frameworks and enabling rigorous, data-driven exploration for potential sex-specific immune factors that may drive distinct outcomes. This study is based on the hypothesis that sex-specific outcomes emerge due to differences in the rates/speeds of select immune components’ responses. The goal of this work is to use dynamic mathematical modeling and Bayesian statistics to identify potential immune regulatory differences between male and female mice after H1N1 infection.
Materials and
Methods: We modified an existing mathematical model of the influenza-induced innate immune response (Ackerman et al., 2022) and fit the model to data from male and female mice infected with PR8 H1N1 influenza (Robinson et al., 2011) to identify sex-specific rates of male and female immunoregulation. In this model, virus causes interferon to be produced. Interferon in turn recruits monocytes. Both interferon and monocytes inhibit viral growth in the model. We implemented a large computational screen to rapidly identify immune rates that may be sex-specific by creating 13 different model scenarios. In each scenario, a different set of model parameters are allowed to take on sex-specific parameter values. We used Bayesian information criteria (BIC) to guide scenario selection because the BIC balances the goodness of fit of the competing models against model complexity. Markov-chain Monte Carlo (MCMC) analysis and global sensitivity analysis of the top model was performed to provide rigorous estimates of the sex-specific parameter distributions and provide insight into which parameters most effect innate immune responses.
Results, Conclusions, and Discussions: The first model scenario considered is the All-Different scenario, where it is assumed that all of the parameters can differ when comparing the immune responses of the influenza-infected males and females. This scenario results in a good model fit, demonstrating that the model can fit the data. The BIC value for this scenario is 72 and is the benchmark BIC value to compare other model scenarios. Another extreme scenario considered is one where all the parameters must share the same value when fitting the male and the female data; this scenario results in a poor fit and a BIC value of 321, significantly larger than the All-Different BIC of 72. Therefore, it is unlikely that immune response rates are the same in males and females. Our computational screen of competing immune regulation scenarios suggests that having sex-specific rates for monocyte induction by interferon and monocyte activity explanation for the difference observed in the male and female responses, based on a significantly improved BIC value of 68. MCMC analysis shows that monocyte induction and monocyte activity parameters have distinct parameter density distributions for male and female values. This model scenario is highly sensitive to monocyte induction for male and female mice, further supporting this parameter as a sex-specific rate. Simulations using this scenario suggest that monocyte activity could be targeted to reduce influenza disease severity in females. Overall, our Bayesian statistical and dynamic modeling approach suggests that monocyte activity and induction parameters are sex-specific and may explain sex-differences in influenza disease immune dynamics. We will use the knowledge gained in this work to build novel mathematical models to aid in identifying additional sex-specific immune rates that could be targeted to improve influenza infection outcomes in females.
Acknowledgements (Optional): This research was funded by NIH R21AI151418 and NSF 1943777.
