Bat lung epithelial cells show greater host species-specific innate resistance than MDCK cells to human and avian influenza viruses
With the recent discovery of novel H17N10 and H18N11 influenza viral RNA in bats and report on high frequency of avian H9 seroconversion in a species of free ranging bats, an important issue to address is the extent bats are susceptible to conventional avian and human influenza A viruses.
To this end, three bat species (Eidolon helvum, Carollia perspicillata and Tadarida brasiliensis) of lung epithelial cells were separately infected with two avian and two human influenza viruses to determine their relative host innate immune resistance to infection.
All three species of bat cells were more resistant than positive control Madin-Darby canine kidney (MDCK) cells to all four influenza viruses. TB1-Lu cells lacked sialic acid α2,6-Gal receptors and were most resistant among the three bat species. Interestingly, avian viruses were relatively more replication permissive in all three bat species of cells than with the use of human viruses which suggest that bats could potentially play a role in the ecology of avian influenza viruses. Chemical inhibition of the JAK-STAT pathway in bat cells had no effect on virus production suggesting that type I interferon signalling is not a major factor in resisting influenza virus infection.
Although all three species of bat cells are relatively more resistant to influenza virus infection than control MDCK cells, they are more permissive to avian than human viruses which suggest that bats could have a contributory role in the ecology of avian influenza viruses.
KeywordsInfluenza Bats Epithelial cells Host resistance Innate immunity Inflammation
- C. perspic
- E. helvum
- LPAI H2N3
low pathogenicity avian influenza H2N3 (A/mallard duck/England/7277/06)
- LPAI H6N1
low pathogenicity avian influenza H6N1 (A/turkey/England/198/09)
- MAA II
Maackia amurensis agglutinins
Madin-Darby canine kidney
- pdm H1N1
pandemic H1N1 2009 (A/California/07/2009)
Sambucus nigra agglutinin
Tadarida brasiliensis lung epithelial cells
- USSR H1N1
human USSR H1N1 (A/USSR/77)
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, 2, 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 , 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 . Murine encephalomyocarditis virus, in contrast, causes severe cytopathic damage in TB1 Lu cells, and Ebola virus shows persistent infection in such cells .
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 . 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 . Furthermore, the chimeric bat virus failed to reassort with conventional influenza viruses in MDCK cells . 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 . 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 . 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 . Similar number of infected cells were found among all seven bat cell lines by immunocytochemical detection of viral nucleoprotein (NP) . 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 . Although there is limited potential for reassortment between human and bat influenza viruses , 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 . 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 .
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.
Bat and MDCK cells
Eidolon helvum (E. helvum) and Carollia perspicillata (C. perspic) cells were generated as described previously . 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 . 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 .
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.
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  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 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.
Bat respiratory epithelial cells were more resistant to influenza virus infection than MDCK cells but were relatively more permissive to avian than human virus strains
At 6 hpi, intranuclear viral NP was extensively detected in MDCK cells infected with each of the four viruses and at 24 hpi NP presence had extended into the cytoplasm (Fig. 1a). Similar extensive pattern of infection was not observed with bat cells. TB1-Lu cells were not readily infected by either human or avian viruses. At 6 hpi, there was little or no detection of NP; at 24 hpi, NP was found only in a limited proportion of nuclei (Fig. 1b). In E. helvum cells infected with LPAI viruses (H2N3 and H6N1), NP was extensively detected by 6 hpi (Fig. 1c). By 24 hpi, NP was detected in the cytoplasm accompanied by widespread detachment of cells. In contrast, in E. helvum cells infected with human viruses (USSR H1N1 and pdm H1N1), NP was only detected in a small proportion of cells even at 24 hpi suggesting that they were less infected with human than avian influenza viruses. A similar infection pattern was observed in C. perspic cells. At 6 hpi with avian viruses, most nuclei of C. perspic cells were strongly labelled for NP and by 24 hpi, cytoplasmic NP was evident with extensive cell loss (Fig. 1d). With human USSR or pdm H1N1 virus, limited number of C. perspic cells were infected even at 24 hpi with NP largely localised to the nuclei. In summary, all three species of bat cells infected with human and avian influenza viruses showed much less viral NP production than correspondingly infected MDCK cells, with TB1-Lu cells producing the least viral NP. Relatively, E. helvum and C. perspic cells showed greater NP production and cell loss at 24 hpi when infected with avian than with human viruses.
Lack of human sialic acid α2,6 linkage receptor contributed to host resistance in TB1-Lu cells
To further explore the functional impact of SA receptors on influenza virus infection, SA receptors were collectively removed from the cell surface using a generic neuraminidase derived from Clostridium perfringens, as previously described  (Fig. 3c). In all cell types, the removal of SA receptors resulted in sharp decrease in virus output (Fig. 3d) indicating their importance in virus infection. However, despite extensive removal of SA receptors, there remained some virus production suggesting the presence of an alternative SA-independent pathway for virus entry. In summary, SAα2,3-Gal and SAα2,6-Gal receptors appear important in bat cells for entry of conventional influenza A viruses, and the deficiency of SAα2,6-Gal receptor in TB1-Lu cells could be an added factor of host resistance to entry of human influenza viruses in this cell type.
Bat cells appeared more viable than MDCK cells infected with influenza viruses
The JAK-STAT pathway did not appear to play a key anti-viral role in bat cells
Protein sequence similarities of JAK1 to 3 between different bat species and other mammalian species (canine and porcine)
Percentage identity to Canis lupus familiaris (%)
Percentage identity to Sus scrofa (%)
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 . 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 . 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, 29, 30]. We speculate that the NF- κB pathway could be functionally more important in bat cells than the JAK/STAT pathway.
In conclusion, all three species of bat cells are more resistant than control MDCK cells to conventional influenza viruses and are relatively more permissive to avian than human viruses which suggest that bats could have a contributory role in the ecology of avian influenza viruses.
We thank Dr. Suresh Kuchipudi for his technical assistance in the detection of sialic acid receptors.
This work was supported by the Biotechnology and Biological Sciences Research Council [grant number BB/J014508/1], a BBSRC DTP awarded to TS.
Availability of data and materials
All data generated or analysed during this study are included in this published article. Contact corresponding author for requests.
Study concept and design: KCC and IE. Experimental prosecution: TS and KCC. Manuscript preparation: TS, KCC and IE. All authors read and approved the final manuscript.
Ethics approval and consent to participate
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.