Viral Interference During Influenza A–SARS-CoV-2 Coinfection of the Human Airway Epithelium and Reversal by Oseltamivir

Abstract

To gain insight into interactions among respiratory viruses, we modeled influenza A virus (IAV)-severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) coinfections using differentiated human airway epithelial cultures. Replicating IAV induced a more robust interferon response than SARS-CoV-2 and suppressed SARS-CoV-2 replication in both sequential and simultaneous infections, whereas SARS-CoV-2 did not enhance host cell defense during influenza infection or suppress IAV replication. Oseltamivir, an antiviral targeting influenza, reduced IAV replication during coinfection but also reduced the host antiviral response and restored SARS-CoV-2 replication. These results demonstrate how perturbations in one viral infection can impact its effect on a coinfecting virus.

Replicating influenza A virus induces a robust interferon response in the human airway epithelium that interferes with SARS-CoV-2 in simultaneous and sequential infections, and both antiviral response and viral interference are reversed by targeting influenza replication with oseltamivir.

The resurgence of seasonal respiratory viruses and continued circulation of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) have focused attention on understanding how respiratory viruses interact in a coexposed host. While intuitively coinfections might be expected to exacerbate disease, increasing evidence suggests that viral interference can also occur, in which infection with one virus decreases replication of the other [1–4]. To explore factors that impact coinfection outcomes, here we probed interactions between 2 high-impact respiratory viruses, influenza A (IAV) and SARS-CoV-2.

One demonstrated mechanism of interference among respiratory viruses is induction of the interferon response, a broad antiviral defense mechanism that is induced by most respiratory viruses, and also suppresses replication of most viruses [5–8]. Within the airway epithelium, the replicating viral genome is the initiating trigger of the interferon response [9]. Viral RNA is sensed by cytosolic innate immune sensors, leading to induction of interferon-stimulated genes (ISGs) and secretion of type I and type III interferons, which signal through interferon receptors to amplify expression of ISGs in both infected and bystander cells. ISGs include >300 genes that block viral replication through diverse mechanisms [10]. While accumulating viral RNA is the signal that initiates interferon and ISG induction, the rate of ISG induction is also influenced by virus-specific mechanisms that antagonize this host defense response [9]. Here, we explored how reducing replication of an interfering virus with an antiviral drug impacts host cell responses and replication of the coinfecting virus, using the air-liquid interface (ALI) culture model of the differentiated human airway epithelium.

METHODS

Primary Human Bronchial Epithelial Cells

Low-passage primary human bronchial epithelial cells (HBEC) from healthy adult donors were obtained commercially (Lonza) and cultured at the ALI according to the manufacturer's instructions (Stem Cell Technologies). Differentiated and mature epithelial cultures displayed beating cilia and mucus production at the time of viral infection.

Generation of Virus Stocks

IAV (H1N1 pdm09, strain A/California/07/2009; American Type Culture Collection [ATCC] VR-1894) was amplified in embryonated eggs and titer of virus stock was determined using plaque assay on MDCK cells (MDCK NBL-2, CCL-34, ATCC) as described previously [5]. SARS-CoV-2 (USA-WA1/2020; BEI Resources) was generously provided by the Wilen laboratory (Yale University, New Haven, CT). To generate virus stocks, SARS-CoV-2 was cultured on Vero E6 cells (CRL-1586; ATCC) and filtered supernatant was used as the virus stock. SARS-CoV-2 titer was determined by plaque assay using Vero E6 cells as described previously [5, 6].

In Vitro Infections

Primary HBEC ALI cultures were inoculated for 1 hour on the apical surface as described previously [5, 6]. For each virus, we used the lowest multiplicity of infection (MOI) for which we observed reproducible exponential replication, which was MOI 0.01 for IAV (H1N1pdm09), and MOI 0.1 for SARS-CoV-2 (USA-WA1/2020). For oseltamivir treatment starting at 16 hours postinoculation, oseltamivir acid (carboxylate, item No. 15779; Cayman Chemical) at a final concentration of 1 μM was added to the basolateral medium at 16, 40, and 64 hours, and cultures were collected at 72 hours to determine viral load and host response. For treatment starting at 40 hours, oseltamivir was added at 40 and 64 hours only, and cultures were collected at 72 hours. All incubations were performed at 35°C to simulate the temperature of the conducting airways.

RT-qPCR of viral RNA and ISG mRNA

For reverse transcription quantitative polymerase chain reaction (RT-qPCR), RNA was isolated from each well of differentiated epithelial cells by incubating each insert with 350 µL lysis buffer at room temperature for 5 minutes, followed by RNA isolation using the Aurum total RNA isolation mini kit (Bio-Rad) and cDNA synthesis using iScript cDNA synthesis kit (Bio-Rad). To quantify IAV viral RNA and mRNA levels for ISGs and the housekeeping gene HPRT, RT-qPCR was performed using SYBR green iTaq universal (Bio-Rad) per the manufacturer's instructions. RT-qPCR for SARS-CoV-2 from cell lysates was performed using the TaqMan assay for the CDC N1 gene primers and probes from Integrated DNA Technologies (catalog No. 10006600) using the Luna Universal Probe One-Step RT-qPCR Kit (New England Biolabs) as previously described [6]. Viral RNA per ng total RNA is represented as fold change from the limit of detection (40 cycles of PCR) as 2^40−Ct. ISG mRNA levels are shown as relative to the level of mRNA for the housekeeping gene HPRT (2^−ΔΔCt).

For RT-qPCR using SYBR green assay, the following primer sets were used: HPRT (forward: 5′-TGG​TCA​GGC​AGT​ATA​ATC​CAA​AG-3′, reverse: 5′-TTT​CAA​ATC​CAA​CAA​AGT​CTG​GC-3′); MX1 (forward: 5′-AGA​GAA​GGT​GAG​AAG​CTG​ATC​C-3′, reverse: 5′-TTC​TTC​CAG​CTC​CTT​CTC​TCT​G-3′); BST2 (forward: 5′-CAC​ACT​GTG​ATG​GCC​CTA​AT-3′, reverse: 5′-TGT​AGT​GAT​CTC​TCC​CTC​AAG​C-3′); RSAD2 (forward: 5′-TCGCTATCTCCTGTGACAGC-3′, reverse: 5′CACCACCTCCTCAGCTTTTG-3′); ISG15 (forward: 5′-CATCTTTGCCAGTACAGGAGC-3′, reverse: 5′-GGGACACCTGGAATTCGTTG-3′); IAV (forward: 5′-GGGTGGACAGGGATGGTAGA-3′, reverse: 5′-TCTGTGTGCTCTTCAGGTCG-3′). BST2 and RSAD2 are shown in the Supplementary data.

Statistical Analysis

Statistical analyses were performed using GraphPad Prism (version 9.3.1). P value for statistical significance of differences between conditions was determined using the Mann-Whitney test and P < .05 was considered statistically significant. Data are shown as mean ±standard error of the mean.

RESULTS

To determine the impact of coinfection on IAV and SARS-CoV-2 replication, we compared single infections to simultaneous and sequential coinfections of IAV (H1N1pdm09) and SARS-CoV-2 (WA/01 strain) at low infectious dose (Figure 1). Infection with IAV 3 days prior to SARS-CoV-2 led to more than 10 000-fold suppression of SARS-CoV-2 replication by 72 hours postinfection (Figure 1B, center). Simultaneous infection with SARS-CoV-2 and IAV also led to a significant decrease in SARS-CoV-2 viral load by day 3 postinfection (Figure 1B, right). Conversely, infection with SARS-CoV-2 3 days prior did not suppress IAV replication (Supplementary Figure 1B, center). Simultaneous coinfection also did not affect IAV replication (Supplementary Figure 1B, right). Compared to SARS-CoV-2 infection alone, expression of the ISGs ISG15 and MX1 was significantly elevated 3 days after SARS-CoV-2 infection in sequential and simultaneous coinfections with IAV and in IAV infection alone (Figure 1C and 1D). However, prior or simultaneous infection with SARS-CoV-2 did not affect ISG induction during IAV infection (Supplementary Figure 1C and 1D). Taken together, these findings show that under these conditions, IAV infection induces more robust ISG expression in the human airway epithelium than SARS-CoV-2, that IAV coinfection enhances the antiviral response during SARS-CoV-2 infection, and that coinfection with IAV restricts SARS-CoV-2 replication to a degree mirroring the degree of enhancement in speed and magnitude of ISG induction in sequential or simultaneous infections.

Figure 1.

Effect of sequential or simultaneous IAV infection on SARS-CoV-2 replication. A, Experimental design of simultaneous or sequential infection experiments in differentiated human airway epithelial cultures. B, SARS-CoV-2 RNA quantification by RT-qPCR on day 1 (24 hours after SARS-CoV-2 infection white bars) and day 3 (72 hours; black bars) represented as fold change from detection limit. C and D, mRNA level of interferon-stimulated genes (C) ISG15 and (D) MX1 by RT-qPCR on day 1 (24 hours after SARS-CoV-2 infection; white bars) and day 3 (72 hours; black bars) relative to mRNA level of housekeeping gene HPRT. Graphs show combined results of 2 independent experiments using primary human bronchial epithelial cultures from different healthy adult donors, each with 4–5 replicates per condition. Mean and standard error of the mean of 9–10 replicates are shown. Mann-Whitney P values are shown for conditions that differ significantly from SARS-CoV-2 infection only on day 3. Abbreviations: Ct, cycle threshold; IAV, influenza A virus; RT-qPCR, reverse transcription quantitative polymerase chain reaction; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.

Because we observed suppression of SARS-CoV-2 replication by IAV but not vice versa, we further probed virus-host-virus interactions by treating cultures with oseltamivir, an influenza virus replication inhibitor, during simultaneous IAV and SARS-CoV-2 coinfection. ALI cultures were infected with IAV, SARS-CoV-2, or both viruses and treated with oseltamivir starting at 16 hours postinfection, based on preliminary experiments to determine a dose and time point that would partially inhibit influenza replication in this model (Figure 2A and Supplementary Figure 3A and 3B). As expected, oseltamivir treatment during IAV single infection or coinfection significantly reduced IAV viral load by day 3 postinfection (Figure 2B). While oseltamivir alone had no effect on SARS-CoV-2 replication (Supplementary Figure 2), oseltamivir treatment rescued SARS-CoV-2 replication during simultaneous infection (Figure 2C). Further analysis of how the timing of oseltamivir impacts IAV replication and host response showed that when administered prior to infection, oseltamivir completely blocked viral replication and ISG induction at 48 hours, showing that viral replication was required for ISG induction (Supplementary Figure 3C–E). When administered at 16 hours postinoculation, oseltamivir significantly decreased both IAV replication and host tissue ISG expression and viral load by 72 hours, but when administered at 40 hours postinfection, treatment did not impact viral replication or ISG levels at 72 hours (Supplementary Figure 4). Taken together, these data confirm that IAV coinfection suppresses SARS-CoV-2 replication and indicate that blocking IAV replication with oseltamivir reduces the IAV-induced host interferon response and rescues SARS-CoV-2 replication by releasing SARS-CoV-2 from IAV-mediated interference.

Figure 2.

Effect of oseltamivir on influenza A and SARS-CoV-2 replication during simultaneous infection. A, Experimental design. B and C, Quantification of IAV (B) and SARS-CoV-2 viral RNA (C) by RT-qPCR at day 3 after single infection or simultaneous coinfection, with or without addition of oseltamivir starting at 16 hours. Viral RNA is expressed as fold change from detection limit. Bars show mean and standard error of the mean of 14 replicates, representing pooled data from 3 independent experiments, each with 4–6 replicates per condition. Conditions with coinfection and oseltamivir are highlighted with shaded bars. Mann-Whitney P values are shown. Abbreviations: IAV, influenza A virus; RT-qPCR, reverse transcription quantitative polymerase chain reaction; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.

DISCUSSION

Circulation of seasonal respiratory viruses including influenza and RSV reached historically low levels during the year of the coronavirus disease 2019 (COVID-19) pandemic, but since 2021 there has been striking reemergence as public health mitigation measures have eased. Influenza viruses and SARS-CoV-2 are anticipated to continue to cocirculate for the foreseeable future. Therefore, a better understanding of the principles and mechanism that govern virus-virus and host-virus interactions during coinfections will be important to interpret clinical observations about how these viruses interact.

IAV and SARS-CoV2 coinfections have been studied in a variety of experimental models and with different virus strains, infectious doses, timing of infections, and a range of outcomes (see Supplementary Materials for extended references). For example, IAV coinfection in the highly SARS-CoV-2–susceptible K18-hACE2 transgenic mouse showed increased disease severity [11]. In contrast, viral interference was seen in the Syrian golden hamster model, with one study showing IAV suppressing SARS-CoV-2 replication, and other showing SARS-CoV-2 suppression of IAV [1, 3]. Studies using differentiated human airway epithelial cultures, like this study, have largely reported interference of influenza with SARS-CoV-2, with the degree of interference varying with virus strains and timing of infections [2, 4, 12–14]. To understand the implications for human coinfections, it will be important to define mechanisms that govern coinfection outcomes in different models and define variables that predict outcomes.

For studies showing interference, the host interferon response appears to be an important mechanism, because the kinetics and magnitude of ISG expression predict interference, and blocking host cell signaling pathways required for this response rescues replication of the suppressed virus [1–4, 12]. Interferon response-dependent interference was also described in previous work from our group and others showing suppression of both IAV and SARS-CoV-2 by prior rhinovirus infection [5–7, 15]. Here we showed that early addition of an IAV replication inhibitor that reduced viral RNA accumulation and downstream induction of the host interferon response also reversed IAV–SARS-CoV-2 interference, providing further support for the idea that robust triggering of host innate antiviral defense is a key mechanism for interference among respiratory viruses.

Infection of differentiated airway epithelial cultures in this study models the earliest stages of infection in the human airway epithelium, a target tissue with robust innate antiviral responses. In this model, IAV replicating from a low infectious dose causes minimal damage to the epithelium that is well controlled by the host innate immune response [5]. While coinfection outcomes are clearly context dependent, we suggest that, in general, viral interference may be more likely to occur when infectious doses are low, allowing a bystander cell interferon response to outpace viral replication [6], or when infections are staggered and the first virus is well controlled by host antiviral defenses at the time of the second exposure, whereas potentiation may be more likely for simultaneous infections with high infectious dose or in a highly susceptible host.

We observed asymmetric interference between IAV and SARS-CoV-2 in this study, with IAV blocking SARS-CoV-2 but not vice versa. This finding is consistent with prior work showing that SARS-CoV-2 has slower ISG induction kinetics than IAV in airway epithelia, likely due to more effective mechanisms of antagonizing the interferon response within infected cells [9]. However, this balance may be different in vivo or for different virus strains (see Supplementary Material for extended references). A limitation of the ALI culture model is the lack of other cell types of the respiratory mucosa, including resident and recruited leukocytes, which can also contribute to the innate immune responses. In a prior study, we examined the nasal interferon response in subjects with mild-to-moderate SARS-CoV-2 infection and we observed that while ISG induction is initially delayed relative to viral replication, once ISGs are induced, there is a robust nasal interferon response lasting up to 4 weeks postinfection [6]. Therefore the ALI culture model may underestimate the potential for SARS-CoV-2 to interfere with subsequent viral infections in human subjects.

Further studies in both animal models and human subjects are needed to fully understand the range of IAV–SARS-CoV-2 coinfection outcomes. However, this study contributes to the recent literature showing the potential for interference between these viruses. In addition, the observation that oseltamivir blockade of IAV replication during coinfection rescues SARS-CoV-2 replication illustrates how, within the same niche, perturbing one virus can alter replication of another through indirect effects on host antiviral defense. These data add to growing evidence that understanding dynamic interactions in the upper respiratory tract niche is fundamental for predicting susceptibility to viral respiratory infection and outcomes of perturbations such as antiviral treatment.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Author contributions. N. R. C. and E. F. F. contributed conceptualization and data visualization, and prepared the original draft. N. R. C., V. T. M., and T. A. W. performed investigations. E. F. F. performed supervision and acquired funding. All authors reviewed and edited the manuscript.

Acknowledgments . We thank the Wilen laboratory for providing the SARS-CoV-2 isolate and VeroE6 cells.

Financial support . This work was supported by the Mercatus Center, George Mason University Fast Grants for COVID-19 from Emergent Ventures (to E. F. F.); the Yale University Department of Laboratory Medicine and COVID-19 Dean’s Fund (to E. F. F.); the Gruber Foundation (fellowship to T. A. W.); and the National Institutes of Health (grant number T32AI007019 to T. A. W.).