Poly I:C enhances susceptibility to secondary pulmonary infections by gram-positive bacteria

Abstract

Secondary bacterial pneumonias are a frequent complication of influenza and other respiratory viral infections, but the mechanisms underlying viral-induced susceptibility to bacterial infections are poorly understood. In particular, it is unclear whether the host's response against the viral infection, independent of the injury caused by the virus, results in impairment of antibacterial host defense. Here, we sought to determine whether the induction of an "antiviral" immune state using various viral recognition receptor ligands was sufficient to result in decreased ability to combat common bacterial pathogens of the lung. Using a mouse model, animals were administered polyinosine-polycytidylic acid (poly I:C) or Toll-like 7 ligand (imiquimod or gardiquimod) intranasally, followed by intratracheal challenge with Streptococcus pneumoniae. We found that animals pre-exposed to poly I:C displayed impaired bacterial clearance and increased mortality. Poly I:C-exposed animals also had decreased ability to clear methicillin-resistant Staphylococcus aureus. Furthermore, we showed that activation of Toll-like receptor (TLR)3 and Retinoic acid inducible gene (RIG-I)/Cardif pathways, which recognize viral nucleic acids in the form of dsRNA, both contribute to poly I:C mediated impairment of bacterial clearance. Finally, we determined that poly I:C administration resulted in significant induction of type I interferons (IFNs), whereas the elimination of type I IFN signaling improved clearance and survival following secondary bacterial pneumonia. Collectively, these results indicate that in the lung, poly I:C administration is sufficient to impair pulmonary host defense against clinically important gram-positive bacterial pathogens, which appears to be mediated by type I IFNs.

Figure 1. Poly I:C increases susceptibility to bacterial infection of the lung.

A. Animals were administered i.n. poly I:C (50 µg), TLR7 agonist (imiquimod 50 µg or gardiquimod 50 µg) , or saline vehicle daily for 2 days. Twenty-four hours after the last dose, animals were given i.t. S. pneumoniae (900 CFU in 30 µl). On day 3, lungs were harvested for enumeration of CFU. *, p<0.05 compared to saline and TLR7 agonist by one-way ANOVA. Data combined from 2 separate experiments. B. Animals were administered i.n. poly I:C (50 µg) or saline vehicle daily for 3 days, followed by i.t. S. aureus (MRSA, LAC USA300; 6×10⁶ CFU) 24 hours later. Lungs were harvested 24 hours after bacterial infection for enumeration of CFU. *, p<0.05 compared to saline group. Data is representative of 2 independent experiments.

Figure 2. Dose-dependent effect of poly I:C on bacterial clearance from the lung.

Animals were administered intranasal poly I:C (pIC, 50 µg) or saline daily for 1 or 3 days. Twenty-four hours after the last dose, animals were given i.t. S. pneumoniae (1×10⁵ CFU in 30 microliters). Forty-eight hours later, lungs were harvested to determine bacterial CFU. *p<0.05, poly I:C (3 doses) versus saline group. Data is combined from 2 separate experiments (n = 7–8/group).

Figure 3. Role of viral pattern recognition pathways in mediating poly I:C-induced immunosuppression.

Age- and sex-matched Cardif ^−/−, Tlr3 ^−/−, Cardif ^−/−/Tlr3 ^−/− and WT animals were administered i.n. poly I:C for 3 days, followed by i.t. S. pneumoniae (1×10⁴ CFU). Animals were sacrificed and lungs collected at 48 hours after bacterial infection for enumeration of CFU. **, p<0.01, WT versus Cardif ^−/−/Tlr3 ^−/− n = 4–7/group. All other comparisons were not significant. Data is representative of 2 separate experiments.

Figure 4. Poly I:C-pretreated animals have impaired bacterial clearance at later timepoints.

Animals were administered i.n. poly I:C (50 µg) or saline daily for 3 days. Twenty-four hours after the last dose, animals were given i.t. S. pneumoniae (1×10⁴ CFU in 30 µl). Lungs were harvested at 1, 3, and 5 days after bacterial challenge to determine bacterial CFU, and the means of each group (n = 4/group) plotted. Error bars represent standard deviations. *p<0.05, poly I:C versus saline group at each timepoint. Data is representative of 2 separate experiments.

Figure 5. Examination of type I IFNs in poly I:C-treated animals.

A. Animals were given i.n. poly I:C (pIC, 50 µg), gardiquimod (100 µg), or saline daily for 3 days. Lungs were collected for analysis of IFN-α4 in whole lung homogenates by ELISA. Mean and SD are graphed, n = 4/group. *p<0.05, poly I:C versus saline group. B. Wildtype (Ifnar ^+/+) and Ifnar ^−/− animals were administered i.n. poly I:C for 3 days, followed by i.t. S. pneumoniae (7×10⁴ CFU). At 48 hours, lungs were collected for enumeration of CFU. n = 5/group. *p<0.05 C. Wildtype animals were administered i.n. poly I:C for 3 days (day 1, 2, and 3), followed by i.t. S. pneumoniae (2×10⁴ CFU) on day 4. Animals were given i.p. anti-IFNAR antibody or Control IgG (1.5 mg on day 1, 0.75 mg on day 3, and 0.75 mg on day 5). Lungs were harvested at 48 hours after S. pneumoniae infection. *p<0.05.

Figure 6. Survival and bacterial clearance at later timepoints in Ifnar ^−/− and wildtype animals following sequential poly I:C treatment and S. pneumoniae pulmonary infection.

A. Age- and sex-matched wildtype (WT) and Ifnar ^−/− animals were administered i.n. poly I:C for 3 days, followed by i.t. S. pneumoniae (1×10⁴ CFU). Animals were monitored for 14 days after pneumococcal infections for indications for euthanasia due to premoribund state. **, p = 0.001 by Log-rank test, poly I:C-treated Ifnar ^−/− vs. poly I:C-treated WT group; n = 7–8/group. All other comparisons were not significant. Data is representative of 2 separate experiments. B, C. In separate experiments, age- and sex-matched wildtype (WT) and Ifnar ^−/− animals were administered i.n. poly I:C for 3 days, followed by i.t. S. pneumoniae (1.5×10⁴ CFU). On day 3 and 4 following bacterial infection, animals were euthanized for harvest of lungs (B) and collection of blood (C) samples, followed by enumeration of bacterial CFU. *, p<0.05; n = 4/group. Data is representative of 2 separate experiments.