RETRACTED: Contribution of Host Immune Responses against Influenza D Virus Infection toward Secondary Bacterial Infection in a Mouse Model

Abstract

Influenza D viruses (IDV) are known to co-circulate with viral and bacterial pathogens in cattle and other ruminants. Currently, there is limited knowledge regarding host responses to IDV infection and whether IDV infection affects host susceptibility to secondary bacterial infections. To begin to address this gap in knowledge, the current study utilized a combination of in vivo and in vitro approaches to evaluate host cellular responses against primary IDV infection and secondary bacterial infection with Staphylococcus aureus (S. aureus). Primary IDV infection in mice did not result in clinical signs of disease and it did not enhance the susceptibility to secondary S. aureus infection. Rather, IDV infection appeared to protect mice from the usual clinical features of secondary bacterial infection, as demonstrated by improved weight loss, survival, and recovery when compared to S. aureus infection alone. We found a notable increase in IFN-β expression following IDV infection while utilizing human alveolar epithelial A549 cells to analyze early anti-viral responses to IDV infection. These results demonstrate for the first time that IDV infection does not increase the susceptibility to secondary bacterial infection with S. aureus, with evidence that anti-viral immune responses during IDV infection might protect the host against these potentially deadly outcomes.

1. Introduction

The 2011 classification of a novel virus as Deltainfluenzavirus, or influenza D virus (IDV), expanded the Orthomyxoviridae family into four genera: influenza A, B, C, and D [1,2]. This virus was first isolated from swine samples that were collected in Oklahoma (D/swine/Oklahoma/1334/2011, OK11), and subsequent bovine serology studies showed that cows are the natural reservoir for IDV [2,3]. Archived sera confirm the presence of IDV in cows since at least 2003 [3,4], and it is speculated to have phylogenetically split from its most similar counterpart, influenza C virus (ICV), around 1900 AD [5,6]. The fact that IDV can co-infect with influenza A and other agglutinating viruses has been speculated as a reason that this virus went undetected until 2011 [3,7]. Additionally, IDV is known to co-circulate with a variety of viruses that cause bovine respiratory disease, which further impeded its isolation [8].

It is suspected that IDV is present in cattle and other small ruminants worldwide [3,8,9,10,11,12,13,14,15,16], but, at the current time, we do not know the level at which IDV could contribute to human infections. Current serology results predict that approximately 1.3% of humans are positive for antibodies against IDV [2], with seropositivity approaching 90% in humans that work closely with cattle [17]. While these results warrant further testing and exploration of IDV, it has been noted that seropositivity does not necessarily indicate that IDV infection occurred [6]. Laboratory experiments confirm that IDV can infect guinea pigs and ferrets, the latter of which is used as a standard animal model to study influenza viruses due to its similar infection pattern to that of humans [2,7,18,19].

It is well established that most influenza-related deaths are due to complications from secondary bacterial infection, including pneumonia [20], and that the host response to the virus can direct susceptibility to these complicated infections [21,22]. Our group and others [21,22,23] have shown that the virus itself can impact the severity of a secondary bacterial infection while using both the viral genes expressed [24,25] and the regulation of host type I IFN expression during primary virus infection [26,27,28]. At this time, little is known regarding the host immune response against IDV infection. Similarly, the impact of IDV infection on susceptibility to secondary bacterial infection has not been examined. In this study, we initiate the characterization of IDV interactions with the host immune response by infecting mice with IDV and evaluating susceptibility to secondary bacterial infection with Staphylococcus aureus (S. aureus). Our work focused on host cellular immune responses that were induced after both primary IDV infection and secondary S. aureus infection using a murine model. We also utilized A549 cells, which are a model cell line for human type II alveolar epithelial cells of the lung that are a major target for infectious microbes [29], to measure type I IFN responses by human cells that were infected with IDV.

Our results demonstrate that IDV infection does not cause clinical symptoms in wildtype mice. Moreover, in response to infection with IDV, we found that macrophage levels are not affected by subsequent secondary bacterial infection. We also determined that IDV infection was protective against clinical signs of secondary bacterial infection, as demonstrated by decreased illness and increased survival in S. aureus-challenged, IDV-infected mice as compared to mice that were inoculated with bacteria alone. When using A549 cells to compare IDV infection with influenza A virus (IAV), which increases host susceptibility to secondary bacterial infections in mice, an effect that is at least, in part, through the downregulation of IFN-β [27], we found that IDV increased A549 cell expression of IFN-β. This study demonstrates, for the first time, that IDV infection does not predispose the murine host to a secondary bacterial infection, and that it can actually improve these potentially deadly outcomes when compared to inoculation with bacteria alone. We will discuss our findings with emphasis on how this new member of the Orthomyxoviridae family compares to current secondary bacterial infection studies with influenza A viruses.

2. Materials and Methods

2.1. Cell Lines

Madin-Darby Canine Kidney (MDCK; American Type Culture Collection, Manassas, VA) cells were maintained in standard MDCK cell growth media prepared while using MEM (Gibco, Carlsbad, CA, USA), 1% MEM vitamin solution (Gibco), 1% antibiotic-antimycotic (Gibco), 1% L-glutamine (Gibco), 5% heat-inactivated FBS (fetal bovine serum) (Atlanta Biologicals, Flowery Branch, GA), 10 mg/mL gentamicin (Gibco), and 3% sodium bicarbonate (Gibco). Human alveolar epithelial cells (A549, ATCC) were maintained in F-12K medium (Gibco) supplemented with 10% FBS (Atlanta Biologicals), 1% antibiotic/antimycotic solution (Gibco), and 10 mg/mL gentamicin (Gibco). Both of the cell lines were kept at 37 °C, 5% CO₂, until infection.

2.2. Virus Preparation

The IDV isolate D/swine/Oklahoma/1334/2011 (OK11) was propagated in 10-day-old embryonated chicken eggs, as previously described [30]. When necessary, the fifty-percent tissue culture infectious dose (TCID₅₀) was determined for the egg-grown stock of OK11, also as previously described [30]. Briefly, the MDCK cell monolayers were washed twice with phosphate-buffered saline (PBS), inoculated with log₁₀ serial dilutions of the diluted stock virus, and incubated for 1 h at 33 °C. The virus was propagated over three days at 33 °C, 5% CO₂ in the presence of MDCK infection media that used 0.3% bovine serum albumin (BSA) (Sigma, St. Louis, MO, USA), instead of FBS, and it was supplemented with 1.0 μg/mL TPCK-trypsin (Worthington Biochemical Co., Lakewood, NJ, USA). The hemagglutination assay was used to confirm virus propagation in individual wells. The OK11 IDV stock had a TCID₅₀ value of 10^6.625 TCID₅₀/mL.

2.3. Mice

Female wild-type (WT) C57BL/6 mice (CD45.2) were purchased from Jackson Laboratories (Bar Harbor, ME, USA) and maintained at the Montana State University (MSU; Bozeman, MT) Animal Resources Center under pathogen-free conditions. All of the mice used in this study were six to eight weeks of age, unless specifically indicated. All care and procedures were in accordance with NIH, USDA and the Guide for the Care and Use of Laboratory Animals (National Research Council, 2011) recommendations. The animal protocols were reviewed and approved by the MSU Institutional Animal Care and Use Committee (IACUC). The Association for Assessment and Accreditation of Laboratory Animal Care accredited MSU (AAALAC; accreditation no. 713).

2.4. Mouse Inoculations, Challenge, Burden, Morbidity, Histology, and Survival

Nonsurgical intratracheal (i.t.) inoculations were performed, as previously described [31]. For OK11 inoculations, the mice were inoculated on day 0 with 100 μL of PBS or with 1.09 × 10⁴ plaque-forming units (PFU) OK11. For the experiments with the LAC strain of S. aureus (methicillin-resistant S. aureus [MRSA] pulsed-field type USA300; a kind gift from Jovanka Voyich at MSU), inoculations of 1.5 × 10⁸ colony-forming units (CFU) were used for challenge on day 7 post-OK11 infection. Our previously described procedure for determining CFU [31] was followed on lung homogenate samples after overnight culturing on tryptic soy agar (TSA) plates. The mice were weighed and monitored for signs of morbidity and mortality daily after inoculation and/or challenge. The lungs used for histological analyses were instilled and fixed in 10% buffered formalin phosphate (Fisher Scientific, Fairlawn, NJ) for 24 h. Paraffin-embedded lung sections (5 μm thick) were stained with hematoxylin and eosin (H&E) and then evaluated under a microscope (Eclipse E800; Nikon Inc., Melville, NY, USA) at 4× and 40× objective magnification.

2.5. Preparation of BALF Samples and Cell Population Analysis

Mice were sacrificed by intraperitoneal (i.p.) administration of 90 mg/kg body weight sodium pentobarbital. Bronchoalveolar lavage fluid (BALF) was obtained by lavaging the lungs with 3 mM EDTA in PBS [32] and the cellular composition was determined by hemocytometer cell counts and differential counts of cytospins after staining with Quick-Diff solution (Siemens; Medical Solutions Diagnostics, Tarrytown, NY, USA).

2.6. A549 Cell Infections

The A549 cells were seeded at 3.0 × 10⁵ cells/mL and then infected 24 h later with either egg-grown IDV (D/swine/Oklahoma/1334/2011, OK11) or IAV (A/Puerto Rico/8/1934, PR8) that had also been propagated in 10-day-old embryonated chicken eggs (10^8.5 TCID₅₀/mL). For RT-qPCR experiments, 75 cm² flasks were infected with a multiplicity of infection (MOI) of 1.0. For ELISA experiments, 24-well plates were infected with a MOI of 0.1. A549 cells that were infected with OK11 were incubated for 1 h at 33 °C, 5% CO₂ before adding MDCK cell infection media containing BSA supplemented with 0.1 μg/mL TPCK-trypsin. PR8-infected A549 cells were treated the same, except the one-hour incubation took place at 37 °C, 5% CO₂.

2.7. RNA Isolation

RNA was isolated from 75cm² flasks of uninfected, PR8-infected, or OK11-infected cells at 24 h post-infection (HPI) using TRIzol™ Reagent (Invitrogen, Carlsbad, CA, USA), as per the manufacturer’s protocol, and then stored at −80 °C until needed.

2.8. One-Step RT-qPCR

RNA from uninfected, PR8-infected, and OK11-infected A549 cells was analyzed for interferon (IFN) RNA transcript while using reverse transcription quantitative-PCR with MultiScribe™ Reverse Transcriptase (Thermo Fisher, Waltham, MA) and the Power SYBR™ Green PCR Master Mix (Thermo Fisher), following the manufacturer’s instructions. RT-qPCR product were detected while using an Applied Biosystems 7300 Real-Time cycler with a program of 30 min at 48 °C for reverse transcription, 10 min at 95 °C for DNA polymerase activation, and 40 cycles of 94 °C for 15 s (denaturing), 60 °C for 60 s (annealing and extension). Gene-specific primers (Eurofins Genomics, LLC, Louisville, KY) were as follows: IFN-β F: 5′-GTCTCCTCCAAATTGCTCTC-3′, R: 5′-ACAGGAGCTTCTGACACTGA-3′; IFN-λ1 F: 5′-GGAGTAGGGCTCAGCGCATA-3′, R: 5′-GCCTCCTCACGCGAGACCTC-3′; IFN-λ2 F: 5′-CGTGGGCTGAGGCTGGATAC-3′, R: 5′-TGGCCCTGACGCTGAAGGTT-3′; IL-27 F: 5′-TGGGCTGAGGCTGGATACAG-3′, R: 5′-TCTGGAGGCCACCGCTGACA-3′; IFN-α2 F: 5′-CCTGATGAAGGAGGACTCCATT-3′, R: 5′-AAAAAGGTGAGCTGGCATACG-3′; and 18S rRNA F: 5′-CTTAGAGGGACAAGTGGCG-3′, R: 5′-GGACATCTAAGGGCATCACA-3′. RT-qPCR data were analyzed using the 2^-ΔΔCt method [33] so that data are normalized to both the uninfected control and a housekeeping gene (18S rRNA) and graphed as relative fold change.

2.9. ELISA

Supernatants were collected from 24-well plates of uninfected, PR8-infected, and OK11-infected A549 cells at 24 and 48 HPI and then centrifuged to remove cell debris for 5 min at 2500 rpm. The cell lysates were then collected using 100μL cell lysis buffer per well. Supernatants and cell lysates were stored at −80 °C until needed. IFN-β protein expression was analyzed while using the Verikine Human IFN Beta ELISA Kit (PBL Assay Science, Piscataway, NJ), as per the manufacturer’s protocol.

2.10. Statistical Analysis

Unless otherwise specified in the figure legends, the reported results are means ± standard deviations (SD) from five mice per group from a single experiment. Each experiment for which the results are presented in this paper was independently performed at least twice with similar results. The differences between the treatment groups were analyzed by analysis of variance (ANOVA) or Student’s t-test (two-tailed) using GraphPad Prism software. Statistical differences with p values of < 0.05 were considered to be significant.

3. Results

3.1. IDV Infection in C57BL/6 Mice is Asymptomatic

The infection of wildtype mice with influenza A and B viruses often results in clinical disease symptoms [34,35]. The intensity of symptoms varies, with body weight loss commonly following increased virus replication. However, primary IDV infection with the OK11 strain did not induce signs of clinical disease in C57BL/6 mice, as evidenced by no weight loss (Figure 1A) or other signs of morbidity (not shown) over the course of infection with OK11 alone. At day 7 following OK11 infection, when the induction of adaptive immune responses typically becomes visible, an increase in the number of neutrophils and lymphocytes being recruited to the lung was found, indicating an active immune response as compared to mock-infected mice (Figure 1B).

Histological examination demonstrates that, when compared to PBS-inoculated mice, the lungs from OK11-infected mice had increased edema and neutrophil infiltration in both the airways and vasculature (Figure 2). This further demonstrates that cellular infiltration and increased inflammation is associated with an active immune response against IDV infection.

3.2. IDV Infection Improves Survival after S. aureus Secondary Infection but Does Not Affect Bacterial Clearance

As one of the major factors affecting influenza-mediated morbidity and mortality, we sought to determine how OK11 infection affects susceptibility to a subsequent bacterial challenge. It is known that mice and humans both experience increased susceptibility to secondary bacterial infections around day 7 after infection with IAV [22,28,36,37,38,39], even when primary influenza infection only results in minor disease. Thus, we next sought to determine whether, despite the absence of clinical disease, OK11 infection could increase host susceptibility to subsequent S. aureus infection. We found that at 24 h after S. aureus infection, mice that were only infected with bacteria had very similar S. aureus lung burden when compared to mice that were infected with S. aureus seven days post-OK11 (Figure 3A). Thus, OK11 primary infection did not alter bacterial load during S. aureus secondary infection when compared to mice that were infected with S. aureus alone.

Neutrophils are a primary responder to respiratory S. aureus infections [20], and defects in neutrophil recruitment and/or bactericidal function have been implicated in reduced bacterial clearance after influenza infection [39,40]. Consistent with the similar S. aureus lung burden detected at 24 h after the infection of either mock- or OK11-infected mice, IDV infection did not alter lung neutrophil recruitment during secondary S. aureus infection, as there were no significant differences in neutrophil populations at 24 h post-bacterial challenge (day 8 post-OK11 infection) (Figure 3B). Interestingly, we found that post-OK11 lung macrophage levels do not decrease in response to S. aureus secondary infection, as they have the same number of macrophages as OK11-only infected mice (Figure 3B). This is in contrast to what we have previously found, where secondary infection of S. aureus following IAV infection or single infection with S. aureus alone leads to a decrease in macrophages and a dominance of neutrophils [24,28].

C57BL/6 (WT) mice challenged with S. aureus on day 7 post-OK11 showed increased inflammation in comparison to mice that were challenged with S. aureus alone (Figure 2). Specifically, histological analysis found that vascular leakage occurs in response to S. aureus at 24 h post-S. aureus challenge, regardless of prior IDV infection, and there is still residual inflammation in the OK11-infected mice following S. aureus challenge (Figure 2 arrowheads).

We did find significantly more macrophages in mice that were infected with OK11 and challenged with S. aureus as compared to mice that were inoculated with S. aureus alone (Figure 3B, p = 0.02). Interestingly, in response to challenge with S. aureus, OK11-infected mice had improved weight loss (Figure 4A), survival (Figure 4B), and recovery from signs of morbidity (Figure 4C) in comparison to mice that were challenged with S. aureus alone. In fact, there was still 100% survival in OK11 + S. aureus-infected mice at day 5 post-S. aureus challenge (day 12 post-OK11 infection) as compared to mice challenged with S. aureus alone that only showed 60% survival (Figure 4B). Together, these results suggest that prior OK11 infection might prime the inflammatory response, ultimately aiding in the clearance of subsequent secondary bacterial infection instead of increasing susceptibility, as found with IAV.

3.3. IDV Enhances IFN-β Expression by Lung Epithelial Cells

As OK11-infected mice were protected from subsequent bacterial challenge, we next sought to determine the general IFN response, as the level and timing of IFNs during influenza infections is known to contribute to the outcome of secondary bacterial infection [26,27,28,41]. We found that infection of A549 cells with OK11 at a MOI of 1.0 resulted in a 7.6 fold increase in IFN-β transcript at 24 h post-OK11 (Figure 5A). Comparatively, at 24 h post-infection, A549 cells that were infected with IAV strain A/Puerto Rico/8/1934 (PR8) at a MOI of 1.0 showed a 7.6 fold increase in IFN-λ2 transcript, a 7.2 fold increase in IL-27 transcript, a 2.5 fold increase in IFN-α2 transcript, and a 2.4 fold increase in IFN-λ1 transcript, but no increase in the IFN-β transcript. No change in IFN-γ transcript was found in the A549 cells that were infected with either virus. We found increased production of IFN-β protein using an ELISA (PBL Assay Science) in both cell lysates (Figure 5B) and supernatants (Figure 5C) from PR8- and OK11-infected A549 cells at 24 and 48 HPI, supporting our findings at the transcript level (Figure 5A).

4. Discussion

Here, we show that the primary infection of mice with IDV does not result in disease, as mice demonstrated none of the clinical symptoms associated with the typical progression of IAV infection. In addition, infection with IDV did not inhibit bacterial clearance after secondary challenge with S. aureus. In fact, we found decreased morbidity and increased survival of IDV-infected mice in response to bacterial challenge when compared to mice that were challenged with bacteria alone. Our findings demonstrate that mice are not susceptible to secondary bacterial infection post-IDV infection and suggest that IDV-mediated anti-viral host responses may help to clear the bacteria by priming a protective inflammatory response. We will discuss these results in the context of IDV pathogenesis and the regulation of secondary bacterial infections as they compare with our previous findings with IAV.

Our results demonstrate that infection of mice with OK11 IDV does not cause mice to exhibit the clinical symptoms that are normally associated with influenza disease progression. Although we did observe the recruitment of lymphocytes and neutrophils to the lung during OK11 infection, we did not observe a decrease in body weight. Usually, the recruitment of inflammatory cells results in increased signs of morbidity, as we have previously found with other IAV subtypes that induce cellular recruitment [24,28]. Additionally, we did not see a decrease in macrophage levels in mice that were infected with IDV, which is a common observation following IAV infection in C57BL/6 mice [42]. These results indicate that the inflammatory environment during IDV infection is subdued when compared to other influenza virus infections. Specifically, the high level of macrophages that are still present at day 7 post-OK11 may aid in preventing the clinical symptoms that were observed after IAV infection.

Similar to other groups [21,22,23], we have previously investigated infection by the PR8 strain of IAV in mice and demonstrated that primary PR8 infection can increase morbidity and mortality of secondary bacterial infection as compared to infection of mice with bacteria alone [25,27,32]. In this study, we demonstrate that there is a protective effect of primary IDV infection during secondary S. aureus infection that is not observed in mice infected with bacteria alone. Specifically, OK11-infected mice challenged with S. aureus were less susceptible to clinical signs of disease (weight loss) and mortality when compared to mice that received S. aureus alone. This suggests that IDV induces host anti-viral mechanisms that are protective against secondary bacterial infection. Interestingly, we found that, in addition to OK11 infection alone, secondary bacterial infection of OK11-infected mice also does not decrease the level of macrophages. This suggests that macrophages may be involved in mediating protection from secondary bacterial challenge. Our previous work with PR8-infected C57BL/6 mice showed that protective alveolar macrophages are depleted over the course of IAV infection and replaced by damaging inflammatory monocytes/neutrophils that contribute to secondary bacterial susceptibility [24,28]. However, Califano, Furuya, and Metzger (2013) demonstrated that macrophage dysfunction, rather than depletion, in C57BL/6 mice that were infected with IAV is a factor that contributes to increased susceptibility to secondary bacterial infection [42], and our previous PR8-infected mouse data supports this. It will be important to determine the macrophage phenotypic properties over the course of IDV infection, and compare them with IAV responses to define how macrophages contribute to this protection from secondary bacterial infection.

Previous work from ours, as well as other groups, has demonstrated that differential regulation of IFNs (type I, II, and III) mediated by the host and/or altered by viral antagonism regulate susceptibility to secondary bacterial infection [26,27,43]. Type I IFN, such as IFN-β, and type III IFNs (IFN-λs) have been shown to be involved in activating interferon-stimulated genes (ISGs) that limit the spread of infection and inhibit viral replication [44,45], as well as have a role in protecting mice from secondary bacterial infection [27]. On the other hand, type II IFN (IFN-γ) has been shown to have a detrimental effect during secondary bacterial infection [22,46]. Here, we evaluated the IFN response in A549 cells that were infected with either PR8 or OK11 to determine the early innate responses of epithelial cells that would be the initial target for infection by IDV. We demonstrate that, in the first 24 h of infection, OK11-infected A549 cells show increased IFN-β transcript, while PR8-infected A549 cells show increases in IFN-λ2 and IL-27, but not IFN-β. This is followed by an increase in IFN-β protein expression in the supernatant of OK11-infected A549 cells at 24 h post-infection and a further increase in both cell lysate and supernatant of OK11-infected A549 cells at 48 h post-infection. This finding suggests that the mechanism for IDV-mediated protection in vivo might be due to prolonged IFN-β production since IFN-β is known to be more potent than type III IFNs and activates more ISGs [45], which we have previously found to be the mechanism for survival after infection with the IAV strain A/swine/Texas/4199-2/98 [28]. To determine whether this protection correlates with IFN-β expression in vivo, we are currently exploring the IFN response in mouse models at various timepoints over the course of IDV and S. aureus infection. We predict that OK11-infected mice will have increased IFN-β expression when compared to bacteria alone due to the previously established protective effect of IFN-β in PR8-infected mice. Our group has also noted that PR8-infected mice show increases in IL-27 expression [28], which is important because IL-27 is another antiviral cytokine that, in response to IAV infection, contributes to increased host susceptibility to secondary bacterial infection [47]. More research investigating the specific contributions of individual cell types during the innate immune response against IDV infection is needed. Investigating ISGs that IDV infection induces will also help in defining the specific contributions that IFN make toward the protective effect we observe in OK11-infected mice that are subsequently infected with S. aureus.

IDV was discovered less than a decade ago, and we have limited understanding of IDV pathogenicity and its potential involvement in secondary bacterial infections. Here we show, for the first time, that IDV provides a potential protective effect against secondary bacterial infections, rather than the detrimental effect that is often associated with IAV. Through our extensive research with IAV, we currently have genetic tools available to better understand IDV pathogenicity in both primary virus infection and secondary bacterial infection. We are continuing to explore IDV infections in both tissue culture and animal models, including understanding how the innate and adaptive arms of the immune system can direct immunity against both primary IDV and IDV-associated secondary bacterial infections.