Severity of neonatal influenza infection is driven by type I interferon and oxidative stress

Abstract

Neonates exhibit increased susceptibility to respiratory viral infections, attributed to inflammation at the developing pulmonary air-blood interface. IFN I are antiviral cytokines critical to control viral replication, but also promote inflammation. Previously, we established a neonatal murine influenza virus (IV) model, which demonstrates increased mortality. Here, we sought to determine the role of IFN I in this increased mortality. We found that three-day-old IFNAR-deficient mice are highly protected from IV-induced mortality. In addition, exposure to IFNβ 24 h post IV infection accelerated death in WT neonatal animals but did not impact adult mortality. In contrast, IFN IIIs are protective to neonatal mice. IFNβ induced an oxidative stress imbalance specifically in primary neonatal IV-infected pulmonary type II epithelial cells (TIIEC), not in adult TIIECs. Moreover, neonates did not have an infection-induced increase in antioxidants, including a key antioxidant, superoxide dismutase 3, as compared to adults. Importantly, antioxidant treatment rescued IV-infected neonatal mice, but had no impact on adult morbidity. We propose that IFN I exacerbate an oxidative stress imbalance in the neonate because of IFN I-induced pulmonary TIIEC ROS production coupled with developmentally regulated, defective antioxidant production in response to IV infection. This age-specific imbalance contributes to mortality after respiratory infections in this vulnerable population.

Figure 1.. Type I interferons are deleterious during neonatal influenza virus infection, while type III interferons are essential to protection.

Three-day old neonatal and 8-week-old adult mice were intranasally infected with PR8 influenza A virus. (a) Wild-type (dashed line) (n=31) and IFNαβR^−/− (solid line) (n=46) neonates were infected and tracked for survival (5 total independent experiments at 2 separate vivarium, 3 experiments performed at Drexel University (DU) and 2 experiments performed at Biomedical Research Foundation Academy of Athens(BRFAA) (b) At DU, 8-week-old wild-type (open circle) (n=12, 3 independent experiments) and IFNαβR^−/− (closed circle) (n=15, 4 independent experiments) mice were intranasally infected with PR8 influenza A virus and weights were tracked. Viral burden was determined at 1-, 3-, and 6-days post-infection by real-time PCR and is reported as the fold change relative to a 1-day post-infection wild type neonatal mouse for (c) IFNαβR^−/− and (e) Ifnlr1^−/− neonatal mice. (d) At BRFAA, wild-type (dashed line) (n=9, 2 separate experiments) and Ifnlr1^−/− (black line) (n=12, 2 independent experiments) and IFNαβR^−/− Ifnlr1^−/− (blue line) (n=14, 2 independent experiments) neonates were infected and tracked for survival. For viral load analysis, protocol 1 was used for IFNαβR^−/− at DU and protocol 2 was used for Ifnlr1^−/− at BRFAA. Statistical differences between wild type and transgenic animal survival were assessed by using log-rank (Mantel-Cox) test and Mann-Whitney test when comparing non-parametric values for viral loads and weight loss, where denoted *p<0.05, ****p<0.0001

Figure 2. Type I interferons accelerate mortality and worsen pathology after neonatal influenza virus infection

(a-b) Three-day-old and 8-week-old wild type mice were intranasally infected with PR8 influenza A virus or saline. Recombinant IFNβ (1000 units for neonates or 4000 units for adults) or saline (sham) was administered 24 hours post-infection per the (a) experimental scheme and (b) survival was tracked. Infected neonates who received IFNβ treatment (black solid line, n=12) are compared to sham-treated infected neonates (black dash line, n=11), uninfected neonates with IFNβ treatment (green dashed line, n=8) and infected adult mice with IFNβ treatment (red dash line, n=6) (2–3 independent experiments). (c-h) In separate experiments, neonatal wild type and IFNαβR^−/− mice were infected at 3 days of age intranasally with PR8 influenza A virus. (c) Pathology severity scores were assessed at 3- and 6-days post-infection (3 independent experiments) (n=5–11). (d) Representative images demonstrate increased alveolitis and peribronchiolitis in wild type animals. Scale bar: 50 μm (High magnification) or 500 μm (Low magnification). Absolute cell counts of specified cell types per 100mg of lung was determined by flow cytometry in the (e and g) bronchoalveolar lavage and (f and h) interstitial lung at 1- (e and f) and 3-days post-infection (g and h) (2 independent experiments, n=4–6). Statistical differences between wild type and IFNβ treatmentsurvival were assessed by using log-rank (Mantel-Cox) test. Student’s T test was used when comparing 2 groups for pathology severity scores and immunophenotyping, where denoted *p<0.05, **p<0.01, ****p<0.0001.

Figure 3. Neonates have increased oxidative stress and reactive oxygen species production following influenza viral infection

To determine oxidative stress and reactive oxygen species production in the infected neonate versus adult, 3-day-old neonatal and 8-week-old adult mice were intranasally infected with PR8 influenza A virus. Mice were harvested 2 days post-infection. (a) Oxidative stress imbalance was determined with CellROX staining after the infected whole lung cell suspension was treated with IFNβ or saline (control), ex vivo for an hour. Average mean fluorescence intensity (MFI) of CellROX in the specified immune cell populations in neonates (white bars) (n=7, 2 independent experiments) and adults (black bars) post IFNβ treatment (n=6, 2 independent experiments). (b) To quantify ROS production in neonates versus adults, dihydroethidium staining was performed. Two days post-infection, the whole lung cell suspension was treated with IFNβ or saline (control), ex vivo for an hour. The number of DHE⁺ CD45⁻ cells post-IFNβ treatment is compared to age-matched saline treated uninfected controls for WT neonates (white bar), IFNαβR^−/− neonates (gray bar), and adults (black bar) (n=5–6 in each group, 2 experiments). Statistical differences between groups were assessed using Mann-Whitney test when comparing 2 groups for MFI and absolute cell count differences, where denoted *p<0.05, **p<0.01.

Figure 4. Influenza-infected neonates have greater oxidative stress imbalance in response to IFNβ

3-day-old neonatal and 8-week-old adult WT mice were intranasally infected with PR8 influenza A virus. Age-matched uninfected mice were used as controls. Mice were harvested 2-days post-infection and Type II epithelial cells were isolated. The TIIEC were treated with an antioxidant N-acetyl cysteine (NAC), IFNβ, or tert-Butyl hydroperoxide (TBHP), or media (Untreated) ex vivo for an hour. NAC treated wells were used as background staining controls. The CellInsight CX7 high content screening platform was used to quantify CellROX intensity at the individual cell level. (a) Representative images at 10X magnification from IFNβ-treated uninfected and infected neonates and IFNβ-treated infected adults are depicted. The TIIEC were co-stained with DAPI (blue), WGA (red) and CellROX (green). (b) The percentage of CellROX positive cells relative to total cells in each treatment group is indicated. (c) Fold change of CellROX intensity relative to age-matched uninfected animals. n=9 in neonates, n=3 in adults, 3 independent experiments. To confirm the role of IFNβ in the neonatal program of oxidative stress imbalance during influenza virus infection, IFNαβR^−/− and WT neonates were infected as above and TIIEC were harvested 2 days post-infection. (d) Relative fluorescence units of average CellROX intensity by flow cytometry in IFNβ-treated WT (white bar) and IFNαβR^−/− neonates (black bar) is shown. Statistical differences between groups were assessed using Student’s T test was used when comparing 2 groups, where denoted *p<0.05, **p<0.01.

Figure 5. Neonates fail to upregulate antioxidants during influenza virus infection

To determine changes in transcription of antioxidant enzymes in the infected animal, 3-day-old WT (black bars) and IFNαβR^−/− (gray bars)neonatal and 8-week-old (white bars) mice were intranasally infected with PR8 influenza A virus. Lungs were harvested at 1- and 3- days post infection, and from age-matched uninfected animals. (a) Sod3, (b) Gpx3, (c) Gss, and (d) Prdx1 transcriptions were normalized to the housekeeping gene gapdh (n=3–6 for each group). Data presented relative to uninfected age matched mice. Data from 3 independent experiments. Statistical differences between groups were assessed using one way ANOVA to compare multiple groups, where denoted ns=non-significant, *p<0.05, **p<0.01.

Figure 6. Antioxidant treatment partially rescues neonates but not adults from influenza-mediated mortality

To determine if exogenous antioxidant could rescue neonatal and adult animals from oxidative stress imbalance-mediated mortality, animals were treated with 150 mg/kg of a pharmaceutical grade antioxidant, N-acetylcysteine (Acetadote) or sham at indicated time points (a, c and e) and tracked for survival (neonates, b and d) or weight loss (adults, f). Treatment with Acetadote partially rescues neonatal mice from influenza mediated mortality when started prophylactically (a and b); Acetadote (dashed line) (n=29), sham (solid line) (n=27) 6 independent experiments. When Acetadote is started post-infection (c and d), there is improved survival; Acetadote (solid line) (n=18), sham (dashed line) (n=9) 4 independent experiments. In contrast, adults given the same weight-adjusted dose of NAC have no improvement in morbidity as demonstrated by similar weight loss kinetics (b and f), NAC (black circle) (n=5, 3 males and 2 females), PBS (open circle) (n=5, 3 males and 2 females), 2 independent experiments. Statistical differences between treated and control animals’ survival on Kaplan Meier survival curve was assessed by using log-rank (Mantel-Cox) test, where denoted *p<0.05, **p<0.01.