Effects of aging on influenza virus infection dynamics

Abstract

The consequences of influenza virus infection are generally more severe in individuals over 65 years of age (the elderly). Immunosenescence enhances the susceptibility to viral infections and renders vaccination less effective. Understanding age-related changes in the immune system is crucial in order to design prophylactic and immunomodulatory strategies to reduce morbidity and mortality in the elderly. Here, we propose different mathematical models to provide a quantitative understanding of the immune strategies in the course of influenza virus infection using experimental data from young and aged mice. Simulation results suggested a central role of CD8(+) T cells for adequate viral clearance kinetics in young and aged mice. Adding the removal of infected cells by natural killer cells did not improve the model fit in either young or aged animals. We separately examined the infection-resistant state of cells promoted by the cytokines alpha/beta interferon (IFN-α/β), IFN-γ, and tumor necrosis factor alpha (TNF-α). The combination of activated CD8(+) T cells with any of the cytokines provided the best fits in young and aged animals. During the first 3 days after infection, the basic reproductive number for aged mice was 1.5-fold lower than that for young mice (P < 0.05).

Importance: The fits of our models to the experimental data suggest that the increased levels of IFN-α/β, IFN-γ, and TNF-α (the "inflammaging" state) promote slower viral growth in aged mice, which consequently limits the stimulation of immune cells and contributes to the reported impaired responses in the elderly. A quantitative understanding of influenza virus pathogenesis and its shift in the elderly is the key contribution of this work.

FIG 1

NK cell and CD8⁺ T cell levels. (a) NK cell infiltration in the lungs of young and aged mice; (b) CD8⁺ T cell infiltration in the lungs of young and aged mice; (c) percentage of activated NK cells (CD3⁻, CD19⁻, and DX5⁺) determined by the expression of the activation marker CD69⁺; (d) percentage of activated CD8⁺ T cells (CD3⁺, CD4⁻, and CD8⁺) determined by the expression of the activation marker CD69⁺. Asterisks indicate statistically significant differences between aged and young animals: *, P < 0.05; **, P < 0.01; ***, P < 0.001. Data are reproduced from reference .

FIG 2

Cytokine levels in the supernatants of lung homogenates. The time course of IFN-α (a), IFN-β (b), IFN-γ (c), and TNF-α (d) concentrations is shown. Asterisks indicate statistically significant differences between aged and young animals: *, P < 0.05; **, P < 0.01; ***, P < 0.001. Data for IFN-γ and TNF are reproduced from reference . Data for IFN-α and IFN-β are newly unpublished.

FIG 3

Schematic representation of the influenza virus infection model and immune strategies. Epithelial cells are considered in one of the following states: susceptible (U), infected (I), and refractory (R). Infected cells produce virus particles (V) that infect other susceptible cells. Activated CD8⁺ T cells (T) and activated NK cells (K) kill infected cells. IFN-α (F_α), IFN-β (F_β), IFN-γ (F_γ), and TNF-α (F_TNF) can promote the state of resistance to infection in epithelial cells.

FIG 4

Model fitting for young and aged mice. Viral titer data from reference and simulation results for young mice (a) and aged mice (b) are shown. Experimental results show that the viral titer was not detectable (less than 50 PFU/ml, shown with a horizontal dashed line) at day 9 for 6 young mice and at day 11 for 6 aged mice. Asterisks indicate statistically significant differences between aged and young animals: *, P < 0.05; ***, P < 0.001.

FIG 5

Cell fraction mean from bootstrap fits of model M7. Further details of the cell distribution can be found in Fig. SC.1 to SC.4 in the supplemental material.

FIG 6

Parameter distribution from bootstrap fits of model M7. Asterisks indicate statistically significant differences between aged and young animals: *, P < 0.05.

FIG 7

Parameter ensembles from bootstrap fit of model M7.

FIG 8

Mean basic reproductive number (R_N) from bootstrap fits of model M7. Further details of the distribution can be found in Fig. SC.4 in the supplemental material. Asterisks indicate statistically significant differences between aged and young animals: *, P < 0.05.