Background
Bats (order Chiroptera) are natural reservoirs for zoonotic viruses that cause some of the deadliest diseases in humans, including filoviruses (such as Ebola and Marburg viruses), lyssaviruses, severe acute respiratory syndrome (SARS)-related coronaviruses and henipaviruses (e.g. Hendra and Nipah viruses) [
1‐
3]. Despite being hosts to such an array of pathogens, bats generally show mild or no clinical symptoms to their presence, a phenomenon that is largely a mystery and a potential biomedical treasure trove that could offer new insights into the treatment and control of such pathogens in humans and affected animals. The lack of illness does not mean that bat cells are not infected by such viruses. Bat cells are susceptible to infections with paramyxoviruses and filoviruses [
4], and show varying degree of permissiveness to virus replication, which is a pre-requisite for the hosts to acquire carrier status. Bat lung epithelial cells (TB1-Lu) of
Tadarida brasiliensis display resistance to reovirus infection; infected cells show no cytopathic effects and rapid decline in virus production; however, low virus release is maintained for at least 2 months [
5]. Murine encephalomyocarditis virus, in contrast, causes severe cytopathic damage in TB1 Lu cells, and Ebola virus shows persistent infection in such cells [
5].
Recently, two novel influenza viruses, H17N10 and H18N11, were identified in bats by deep sequencing analyses (although live viruses have not been directly isolated) which have understandably caused much speculation about their zoonotic potential [
6]. These viruses are, however, highly divergent from conventional mammalian and avian influenza A viruses. Chimeric virus housing the six core genes from bat H17N10 virus replicated well in human primary airway epithelial cells and mice, but poorly in avian cells and chicken embryos without further adaptation [
7]. Furthermore, the chimeric bat virus failed to reassort with conventional influenza viruses in MDCK cells [
7]. Bat viral ribonucleopolymerase (vRNP) complex subunits (PB1, PB1 and PA) were not functionally interchangeable with corresponding human virus-derived vRNP subunits suggesting there is limited reassortment potential between bat and human influenza viruses [
8]. However, vRNP from bat H17N10 virus is able to drive with high efficiency the non-coding region of human H1N1 virus (A/WSN/1933) in vRNP minigenome reporter assays, highlighting the possibility of viable reassortment between bat and human influenza viruses [
9]. Although the issue of functional reassortment between native bat and conventional influenza A viruses has not been fully resolved, its likelihood is presently considered low.
Single-cycle green fluorescent protein (GFP) reporter virus (human A/WSN/33) was variably able to infect all eleven bat cell lines, derived from seven bat species [
8]. Similar number of infected cells were found among all seven bat cell lines by immunocytochemical detection of viral nucleoprotein (NP) [
4]. Human virus-derived vRNP complex was shown to perform better than avian virus-derived vRNP complex in the same A/WSN/33 viral backbone at progeny virus release, based mostly on the use of TB1-Lu bat cells, which appear inherently resistant to influenza virus infection [
8]. Although there is limited potential for reassortment between human and bat influenza viruses [
8],
Pteropus alecto kidney cells were able to produce reassorted progeny from human H1N1 (A/WSN/1933) and highly pathogenic avian influenza (HPAI) H5N1 (A/Vietnam/1203/04) viruses [
10]. Collectively, these findings appear to indicate that bat cells are susceptible to infection with conventional mammalian and avian influenza viruses. However, we are unclear about the relative permissiveness of bat respiratory epithelial cells to conventional influenza viruses in the production of viable progeny. Although bats are not known to act as hosts for human and avian influenza viruses, the potential epidemiological significance of avian influenza virus infection in bats was highlighted by the recent discovery that around 30 out of 100 free ranging
Eidolon helvum (fruit bats) in Ghana were serologically positive for avian H9 virus [
11].
We report here on the relative susceptibility of lung epithelial cells from three diverse bat species, T. brasiliensis (a medium insectivorous bat), E. helvum, (a large fruit bat) and C. perspicillata (a small mainly fruit, and insect eating bat), to avian and human influenza A viruses. We found that all three species of bat cells were more resistant than control Mardin-Darby canine kidney (MDCK) cells, in terms of reduced progeny virus production and higher cell viability, which appeared not to depend on JAK/STAT signalling. Although the three species of bat cells showed variation in resistance to infection, they were relatively more permissive to avian than human influenza viruses which could be important in the ecology of avian influenza viruses.
Methods
Bat and MDCK cells
Eidolon helvum (
E. helvum) and
Carollia perspicillata (C. perspic) cells were generated as described previously [
12]. MDCK (ATCC CCL-34), TB1-Lu (ATCC CCL-88),
E. helvum and C. perspic cells were cultured in DMEM-Glutamax I (high glucose) (Life Technologies) supplemented with 10% foetal calf serum and 1% penicillin streptomycin (P/S).
Virus infection and detection
Human USSR H1N1 virus (A/USSR/77) (USSR H1N1), pandemic H1N1 2009 virus (A/California/07/2009) (pdm H1N1), low pathogenicity avian influenza (LPAI) H2N3 virus (A/mallard duck/England/7277/06), and LPAI H6N1 virus (A/turkey/England/198/09) were used. Viruses were propagated in 10-day old embryonated chicken eggs in accordance to Operation of the Animals (Scientific Procedures) Act 1986 (UK). Forty-eight hours post-infection (hpi), allantoic fluid was harvested and virus was titrated and stored at − 80 °C. Cells were washed once with phosphate-buffered saline (PBS) and infected with specified dose of virus in serum-free infection medium (Ultraculture, Lonza) supplemented with 1% P/S, 1% glutamine and 500 ng/ml tosyl phenylalanyl chloromethyl ketone (TPCK) trypsin. After 2 h of virus incubation, cells were washed three times with PBS, and incubated in fresh infection medium for a further specified period. For virus quantification, focus forming assay was performed on MDCK cells that were infected for 6 h. Cells were immunolabelled using an EnVision+ system-HRP (DAB) kit (Dako) according to the manufacturer’s instructions. Mouse monoclonal antibody (AA5H; Abcam) was used at 1 μg/ml for 40 min for viral nucleoprotein (NP) detection. Positively stained cells were visualised and counted using an inverted microscope. The average number of cells positive for NP in six wells of a 96-well plate was used to calculate infectious focus-forming units (ffu) of virus per microlitre of infection volume from which MOI dosage was derived. A virus dose of 1.0 MOI is regarded as the minimum volume of virus needed to infect each MDCK cell in a culture well as determined by NP detection at 6 hpi. All virus work was conducted in BSL-2 containment.
Quantification of viral M-gene RNA expression
Total RNA was isolated using an RNeasy Plus minikit (Qiagen) and cDNA synthesis reaction was performed with 1 μg of total RNA using a SuperScript III first-strand synthesis kit (Invitrogen). Viral M-gene RNA expression was quantified by TaqMan real-time PCR as previously described [
13,
14]. Amplification was carried out in triplicates from three biological replicates. The conditions for PCR were initial denaturation at 95 °C for 10 min followed by 45 cycles of 95 °C for 15 s, 55 °C for 30 s and 72 °C for 1 s and final cooling to 4 °C. RNA expression levels were normalised to the 18S rRNA gene.
Host influenza virus receptors
Cells were grown on 8-well Lab-Tek II chamber slides (Nunc). Lectin labelling was performed as previously described [
15]. Cells were fixed with 4% paraformaldehyde for 10 min at room temperature, after which endogenous biotin activity was blocked using a streptavidin/biotin blocking kit (Vector Laboratories). Cells were incubated with fluorescein isothiocyanate (FITC)-labelled
Sambucus nigra agglutinin (SNA), specific for human influenza receptor type SAα2,6-Gal, and biotinylated
Maackia amurensis agglutinin II (MAA II) (Vector Laboratories) specific for avian influenza receptor type SAα2,3-Gal overnight at 4 °C, at 10 μg/ml each. After overnight incubation, cells were washed twice with Tris-buffered saline (TBS) and subsequently incubated with streptavidin-Alexa Fluor 594 conjugate (Invitrogen) at room temperature for 2 h. Finally, cells were washed three times with TBS and mounted with ProLong Gold antifade reagent with 4′,6′ diamidone-2-phenylindole (DAPI) (Invitrogen). Neuraminidase derived from
Clostridum perfringens (11,585,886,001; Roche) was used at 0.05 U/ml in culture medium for 4 h at 37 °C for the collective removal of SA receptors [
16].
Endosomal uptake of siRNA
Cells were transfected using the Viromer Blue system (Lipocalyx) with a SignalSilence Control siRNA (Fluorescein Conjugate) (#6201, Cell Signaling Technology) according to manufacturer’s instructions.
Flow cytometry for quantification of cell viability
Cell viability, based on entry of fluorescent dye into cells with compromised cell membranes, was quantified using a LIVE/DEAD Fixable far red fluorescent kit (L10120; Life Technologies) in a BD FACS CANTO II flow cytometer (BD Biosciences). Data analysis was carried out using the Kaluza Analysis software (Beckman Coulter). A heat-killed control, subjected to 60 °C for 20 min, and uninfected control were used to determine the fluorescence threshold between viable and dead cells.
Western blotting
Radioimmunoprecipitation assay (RIPA) buffer (Santa Cruz) supplemented with 1% phenylmethylsulfonyl fluoride (PMSF) (Santa Cruz), 1% cocktail inhibitor and 1% sodium orthovanadate (Santa Cruz) was used to lyse cells. Bio-RAD protein assay was used to determine protein concentration (Bio-Rad). Primary antibodies used were mouse anti-viral NP (at 1:3000 dilution; PA5–32242, Pierce), goat anti-viral PB1 (at 1:10000 dilution; 17,601, Santa Cruz), goat anti-viral M1 (at 1:2000 dilution; ab20910, Abcam), mouse anti-β-actin (at 1:10000 dilution; A5316, Sigma) and secondary antibodies used were donkey anti-goat IgG (at 1:10000 dilution; sc-2020, Santa Cruz) and goat anti-mouse IgG (at 1:1000 dilution; HAF007, R&D Systems).
Inhibitors of JAK-STATsignalling
Pyridone 6 (Merck), a JAK inhibitor, was applied at 5 μM [
17] to cells for 20 h prior to infection at 37 °C. Cells were rinsed twice with PBS and then infected with USSR H1N1 virus at 1.0 MOI. DMSO treated cells were infected as controls. After 2 h infection, cells were rinsed three times with PBS and fresh infection medium was added with corresponding inhibitor and incubated for a further 22 h before virus titration on MDCK cells using spun supernatants from infected cells.
Statistical analysis
Statistical analysis was performed using GraphPad Prism 6 (GraphPad Software). Student’s t-test, one-way ANOVA and two-way ANOVA were used as appropriate. P values < 0.05 were considered to be significant.
Discussion
In this study, we attempted to address the issue of relative susceptibility of bats to conventional human and avian influenza A viruses which could have fundamental and epidemiological importance given their widespread global distribution. The recent finding of 30% of free ranging
E. helvum (large fruit bats) in Ghana to be serologically positive for avian H9 virus highlights the possibility of transmission of avian influenza viruses into bats [
11]. We found that the lung epithelial cells of three diverse bat species,
T. brasiliensis (a medium insectivorous bat),
E. helvum and
C. perspicillata (a small mainly fruit bat) were consistently more resistant to avian and human influenza A viruses than correspondingly infected control MDCK cells in terms of reduced progeny virus production, less infected cells and greater cell viability. Interestingly, between avian (H2N3 and H6N1) and human (USSR H1N1 and pdm H1N1) viruses, bat cells, in particular
E. helvum and C. perspic, infected with avian viruses were more permissive to virus production and showed greater cytopathogenicity than those infected with human viruses.
There was variation in resistance to influenza virus infection between the three species of bat cells. TB1-Lu cells were most resistant among the three species, producing the least number of infected cells and progeny viruses. Both host SAα2,3-Gal and SAα2,6-Gal receptors appear important in all bat cells for entry of conventional influenza A viruses in that their removal by sialidase treatment led to significant reduction in virus output. The weak presence of SAα2,6-Gal receptor in TB1-Lu cells could account in part for host resistance to the entry of human influenza viruses. There were differences in the expression of viral proteins (PB1, M1 and NP) between the three bat cell types separately infected with avian H2N3 and human USSR H1N1 virus. Such differences in viral protein expression could not readily account for the differences in virus output or cell viability observed between the species of bat cells but hint at the possible presence of translational/post-translational mechanisms of virus inhibition in bat cells as added layers of host innate resistance.
Given that bats play hosts to a diverse range of deadly zoonotic viruses often without serious clinical consequences to themselves, it is reasonable to assume that there are generic innate immune responses in bats that are effective across different viral pathogens. Insights into such processes could provide invaluable basic knowledge for the control and treatment of lethal human infections transmitted by bats. The JAK-STAT pathway is a major signalling pathway of type I interferons and cytokines in the transcriptional activation of anti-viral responses. Based on the use of pyridone 6 (JAK inhibitor), we found that the JAK-STAT pathway in the three species of bat cells appeared not to be critical in the control of influenza virus production. This observation is consistent with the recent finding that certain bat cells have a dampened interferon response due to the replacement of the highly conserved serine residue (S358) in STING, an essential adaptor protein in multiple DNA sensing pathways [
25]. Additional work is needed to further assess (chemically and genetically) the anti-viral role of JAK/STAT signalling in bat cells. The unavailability of bat species-specific reagents is currently a major research bottleneck hampering progress in the area. There are obvious needs for bat species-specific gene sequences and antibodies to be able to conduct quantitative PCR and Western blotting to detect members of the innate immunity such as specific interferons, cytokines and their corresponding responsive gene products. The NF-κB pathway has a complex relationship with influenza A virus. Inhibition of NF-κB signalling in murine and human respiratory epithelial cells has been shown to reduce both virus replication and production of pro-inflammatory cytokines following avian and human influenza virus infections [
26,
27]. However, for the wealth of evidence that supports NF-κB as a pro-viral agent in the promotion of influenza virus propagation, there is credible evidence to show that NF-κB is also a mediator of inflammatory and anti-viral responses [
28‐
30]. We speculate that the NF- κB pathway could be functionally more important in bat cells than the JAK/STAT pathway.