Interferon-Dependent and Respiratory Virus-Specific Interference in Dual Infections of Airway Epithelia

Abstract

Many respiratory viruses cocirculate in the population and multiple infections are commonly reported. The clinical impact of coinfection is unclear and may vary depending on the viral couples involved. Using three-dimensional reconstituted human airway epithelia and clinical viral strains, we investigated the interaction between influenza virus (Flu), respiratory syncytial virus (RSV) and rhinovirus (RV). We showed that Flu and RSV interfere with RV replication, whereas RV does not interfere with either of these viruses. We then experimentally demonstrated that, when present, the interference is not related to a block of viral entry but rather to type I and type III interferon (IFN), the front-line antiviral defense of the respiratory mucosa. Consistent with this observation, we highlighted the differential sensitivity of each virus to IFNs, with RV being the only virus significantly inhibited by IFN-λ and the most sensitive to IFN-α. Finally, as type III IFN is of therapeutic interest due to its low proinflammatory profile, we also assessed and confirmed an inhibitory effect of IFN-λ in the context of persistent RV infections. The present work provides mechanistic clues concerning innate immunity involvement during respiratory virus interactions and confirms that IFN-λ is a promising candidate in the treatment of RV infections.

Figure 1

Change in viral replication in dual versus single infections of reconstituted human airway epithelia. Each virus was inoculated alone or in combination, at the same time or two days after the first virus. For each condition, the log fold change (FC) in apically released virus (measured by RT-qPCR five days post infection) in dual versus single infection is indicated on the Y-axis. The analyzed viral couple is specified on the top of each panel while the chronology of infection is shown on the X-axis (‘then’, after two days; ‘and’, at the same time). Statistical significance relative to single infection was calculated using one-way ANOVA (****P < 0.0001, *P < 0.05). Data are expressed as mean and SEM of at least three replicates.

Figure 2

Change in cell-associated virus in dual versus single infections of reconstituted human airway epithelia. Each virus was inoculated alone or in combination, at the same time or two days after the first virus. For each condition, the log fold change (FC) in cell-associated virus (measured by RT-qPCR four hours post inoculation) in dual versus single infection is indicated on the Y-axis. The analyzed viral couple is specified on the top of each panel, while the chronology of infection is shown on the X-axis (‘then’, after two days; ‘and’, at the same time). Statistical significance relative to single infection was calculated using one-way ANOVA. Data are expressed as mean and SEM of at least at least three replicates.

Figure 3

The presence of type I and type III IFNs in RSV-infected medium interferes with RV replication. A Basal medium (BM) from tissues infected for two days with RSV-A (BM RSV-A) or OC43 (BM OC43) or from non-infected control (BM Ctrl) were transferred to new tissues for 24 hours before infection with RV-A16. For each condition, the log FC in apically released RV-A16 (measured by RT-qPCR three days post infection) relative to BM Ctrl-treated tissues is indicated on the Y-axis. B Same as a but in absence (black bars) or in presence of anti-type I (dark grey bars) or anti-type III (light grey bars) IFN-neutralizing antibodies. For each condition, the log FC in apically released RV-A16 (measured by RT-qPCR three days post infection) relative to BM RSV or BM OC43-treated tissues in the absence of neutralizing antibodies is indicated on the Y-axis. Statistical significance relative to Ctrls was calculated using one-way ANOVA (*P < 0.05). Data are expressed as mean and SEM of two replicates.

Figure 4

Different susceptibility profiles of respiratory viruses to IFN-λ in reconstituted human airway epithelia. Tissues were treated with IFN-λ or IFN-α before (24 h) and during infection with RV-A16, RSV-A and H1N1. For each condition, the log fold change (FC) in apically released virus (measured by RT-qPCR three days post infection) relative to untreated controls (ctrl) is indicated on the Y-axis. Statistical significance relative to the untreated control was calculated using two-way ANOVA (****P < 0.0001, ***P < 0.001, **P < 0.01). Data are expressed as mean and SEM of at least two replicates.

Figure 5

IFN-λ inhibits RV-A16 replication in the context of persistent infections. Infections of airway epithelia were conducted during 25 days in absence or in presence of continuous IFN-λ administration and viral load was quantified using RT-qPCR in apical samples collected at the indicated time points. The statistical significance was calculated with t-tests on the area under the curve (**P < 0.01). Data are expressed as mean and SEM of five replicates.