1. Introduction
Bats are the source of multiple zoonotic viruses including severe acute respiratory syndrome (SARS) coronavirus [
1], Hendra virus [
2], Nipah virus [
3], and Marburg virus [
4]. In fact, evidence suggests that bats host a greater proportion of zoonotic viruses than any other mammalian order [
5], highlighting the importance of identifying novel viruses in bats. Australian pteropid bats are becoming more urbanized and fewer bats are migrating, resulting in a greater chance of contact between bats and humans or domestic animals [
6,
7]. This increased potential for exposure of non-reservoir hosts to bat-borne viruses leads to the increased probability of infection spillovers occurring [
8,
9].
Isolation and phenotypic characterization should be critical components of virus discovery programs because the analysis of novel viral sequences is not currently enough to predict the likelihood of that virus causing a zoonotic disease event [
10]. The likelihood of viral emergence and sustained human-human transmission is influenced by many factors. In addition to environmental factors and host behaviors, specific viral traits and host-pathogen interactions play important roles. For example, low viral pathogenicity resulting in low host mortality influences opportunities for sustained viral transmission; viral tissue tropism and host immune responses determine shedding at sites relevant to transmission; and the establishment of chronic or latent infection may allow for sustained or recurrent viral shedding [
11].
Paramyxoviridae is a family of negative strand RNA viruses currently comprising seven genera—
Rubulavirus,
Henipavirus,
Respirovirus,
Morbillivirus,
Ferlavirus,
Aquaparamyxovirus and
Avulavirus [
12]. PCR and virus isolation have been used to identify many paramyxoviruses in bats globally, in particular, henipaviruses and rubulaviruses [
13,
14]. The genus
Rubulavirus contains the human pathogens parainfluenza virus 2 (hPIV2) and mumps virus (MuV), as well as bat-borne viruses such as Mapuera and Menangle viruses (MapV and MenPV). Viruses within this genus have a cell attachment glycoprotein with neuraminidase and haemagglutinin capability [
15]. In addition to the cell attachment glycoprotein (HN), the rubulavirus genome also encodes a nucleoprotein (N), phosphoprotein (P), V protein, matrix (M) protein, fusion (F) protein and a large polymerase subunit (L) [
16]. The unedited P gene transcript encodes the V protein, whereas the addition of two non-templated G residues by co-transcriptional stuttering of the RNA-dependent RNA polymerase is required for the expression of the phosphoprotein [
16]. MuV and parainfluenza virus 5 (PIV5) also express a short hydrophobic (SH) protein that has been associated with blockage of the TNFα-mediated apoptosis pathway [
17].
PIV5 is most well known as one of the causative agents of Canine Infectious Respiratory Disease Complex (CIRDC), where infection results in self-limiting tracheobronchitis that resolves in 6–14 days when in the absence of any co-infections [
18]. Since the discovery of PIV5 in monkey kidney-cell culture in 1954 [
19], it has been isolated from a wide range of host species including pigs and cattle [
20,
21].
Here we describe the isolation of a novel rubulavirus that we have called Alston virus (AlsPV). AlsPV is closely related to PIV5 and was isolated from pteropid bat urine collected in Alstonville, New South Wales in 2011. This is the first isolation of this novel virus. This paper describes the characterisation of this virus in order to confirm its classification as a rubulavirus, as well as to determine its pathogenic potential.
2. Materials and Methods
2.1. Cell Culture
Cell lines used in the characterization of AlsPV were African Green Monkey (Vero) cells (ATCC), primary
Pteropus alecto kidney (PaKi) cells [
22], Madin–Darby Canine Kidney (MDCK) cells (CSL Ltd., Melbourne, Australia), Madin–Darby Bovine Kidney (MDBK) cells (ATCC), porcine kidney (PK15a) cells (National Animal Disease Centre, Ames, IA, USA) and human cervical (HeLa) cells (ATCC).
With the exception of PaKi cells, all other cell lines were grown in Dulbecco’s Modified Eagle’s Medium (DMEM, Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS, Gibco), 100 units/mL penicillin, 100 μg/mL streptomycin and 0.25 μg/mL amphotericin B (Antibiotic-Antimycotic, Gibco), and 7.5 mM HEPES (Gibco). PaKi cells were grown in Ham’s F12 Nutrient Mixture (Gibco) supplemented as above for normal cell culture media.
2.2. Virus Isolation
For isolations, cells were cultured in Ham’s F12 Nutrient Mixture (Gibco), supplemented as above except for the Antibiotic-Antimycotic that was added at double the normal strength. Virus isolations were conducted using pooled bat urine collected from pteropid bat colonies in Alstonville, New South Wales on the 12 July 2011 and 3 August 2011. Urine collection was conducted as previously described [
23]. Urine was clarified, diluted, and incubated with confluent Vero or PaKi cell monolayers as previously described [
24]. Cell monolayers were observed for at least one week for evidence of virus-induced cytopathic effect (CPE). Supernatants were further passaged onto fresh Vero and PaKi cell monolayers weekly for another two weeks and observed for signs of CPE.
Isolated paramyxoviruses were initially identified using hemi-nested PCR with degenerate primers following the protocols described previously [
25], followed by Sanger sequencing of the PCR products.
2.3. Viruses
In addition to Alston virus, viruses used for in vitro analysis included Teviot virus/Bat/2011/Alstonville (TevPV), porcine rubulavirus (PorV), MapV, Tioman virus (TioPV), MenPV and Hendra virus (HeV). The following reagents were obtained through BEI Resources, NIAID, NIH: hPIV2, Greer, NR-3229; and PIV5, 21005-2WR (Tissue Culture Adapted), NR-42515; and MuV, Enders, NR-3846.
The GenBank accession number for the Alston virus sequence is MH972568. Virus sequences used in phylogenetic analysis TevPV (KP271123), TioPV (NP665871), MenPV (AFY09794), PIV5 (YP138518), hPIV2 (X57559), Achimota virus 1 (AchPV1, JX051319), Achimota virus 2 (AchPV2, AFX75118), human parainfluenza virus 4 (hPIV4, AB543336), bat mumps virus (bat-MuV, HQ660095), MapV (EF095490), MuV (NP054714), simian virus 41 (SV41, X64275), PorV (BK005918), Tuhoko virus 1 (ThkPV1, ADI80715), Tuhoko virus 2 (ThkPV2, GU128081), Tuhoko virus 3 (ThkPV3, GU128082), Sosuga virus (SosPV, AHH02041), and HeV (NP047113).
2.4. Parainfluenza Virus 5 Sequences
Parainfluenza virus 5 strains used in the analysis of AlsPV included 1168 (KC237064), ZJQ-221 (KX100034), SER (JQ743328), BC14 (KM067467), CC-14 (KP893891), W3A (JQ743318), KNU-11 (KC852177), AGS (KX060176), CPI- (JQ743320), CPI+ (JQ743321), 78524 (JQ743319), H221 (JQ743323), 08-1990 (KC237063), D277 (KC237065), DEN (JQ743322), LN (JQ743324), RQ (JQ743327), MEL (JQ743325), and MIL (JQ743326).
2.5. Sequencing
2.5.1. Whole-Genome Sequencing
Supernatant of AlsPV infected Vero cells was prepared for sequencing by ultracentrifugation through a 20% sucrose cushion at 35,000 rpm for 2 h at 4 °C. Total RNA was extracted from the resulting pellet using a Direct-zol RNA Miniprep kit (Zymo, Irvine, CA, USA), including an in column DNaseI digestion, and purified by an RNA Clean and Concentrator kit (Zymo). A REPLI-g WTA Single Cell kit (Qiagen, Venlo, The Netherlands) was utilized for isothermal amplification, followed by processing with a Genomic DNA and Concentrator 10 kit (Zymo). Fragmentation and dual-index library preparation were conducted using Nextera XT DNA Library Preparation kit (Illumina, San Diego, CA, USA), and denatured libraries were sequenced using a 300-cycle MiSeq Reagent kit v2 (Illumina). 100,000 paired-end reads were imported into the VirAMP Galaxy pipeline, trimmed, and assembled using the SPAdes de novo assembly algorithm [
26,
27]. The genome sequence was iteratively extended by mapping trimmed reads back to the parainfluenza virus 5 genome (NC_006430). Genome ends and regions of high variability were confirmed by Sanger sequencing.
2.5.2. Confirmation of Genome Termini
Ligation of genome ends was used to enable sequencing of the 3′ terminus, adapted from a protocol previously developed for influenza virus sequencing [
28]. Genome ends were ligated overnight at 16 °C using 20 U T4 RNA Ligase, 20 U RNasin, 50 μM ATP, 10% PEG8000 and T4 RNA ligase reaction buffer (NEB, Ipswich, MA, USA). Ligation was followed by hemi-nested PCR amplification using a Superscript III One-Step RT-PCR System with Platinum Taq DNA Polymerase (Invitrogen, Carlsbad, CA, USA), then an Expand High Fidelity PCR System (Roche, Basel, Switzerland).
The rapid amplification of cDNA ends (RACE) [
29] was required to determine the 5′ terminus, with some adaptations made to the original method. Briefly, viral RNA was reverse transcribed using a virus specific primer and the Superscript III First-Strand Synthesis Supermix (Invitrogen, Carlsbad, CA, USA). Viral cDNA was RNase H digested, followed by processing with a NucleoSpin PCR Clean-up and Gel Extraction kit (Macherey Nagel, Düren, Germany). Viral cDNA was ligated to an oligonucleotide adaptor (5′-GAAGAGAAGGTGGAAATGGCGTTTTGG-3′) overnight at 16 °C using T4 RNA Ligase (NEB) and amplified by hemi-nested PCR using Platinum Taq DNA Polymerase High Fidelity system (Invitrogen) with virus specific primers and an adaptor specific primer. Fragments of the correct size were purified before sequencing by standard Sanger methods.
2.5.3. Amplicon Sequencing
Amplicon sequencing of the RNA editing site within the P gene was conducted on RNA extracted from Vero cells infected with AlsPV for 72 h in triplicate. RNA was reverse transcribed using oligo(dT) primers with Superscript III Reverse Transcriptase (Invitrogen). Fragments containing the RNA editing site were amplified using an Expand High Fidelity PCR System with primers containing Nextera adaptors, (5′-TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCCAACCCTCTACTTGGCTTGGATTC-3′) and (5′-GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGCCGGGTATCCATCCCTCTCACTG-3′). Controls were included in triplicate to account for the mutation rate of the reverse transcriptase and the polymerase. PCR controls were produced by amplifying pCAGGS constructs containing non-edited fragments with an Expand High Fidelity PCR System. Reverse transcription controls were produced by transfecting constructs containing non-edited fragments into Vero cells for 24 h using Lipofectamine LTX Reagent (Invitrogen), followed by total RNA extraction, reverse transcription and PCR amplification. All PCR products were amplified with Nextera adaptor-specific primers using a HiFi HotStart ReadyMix PCR system (Kapa Biosystems, Wilmington, MA, USA). The DNA library was sequenced using a 600-cycle MiSeq Reagent Kit v3 (Illumina). Data were analyzed using CLC Genomics Workbench 8.5.1 (Qiagen) basic variant detection tool with a minimum frequency output of 0.05%. The prevalence of editing was standardized per 100,000 reads.
2.6. Virus Quantification
10-fold serial dilutions of virus stocks were combined with Vero cells in 96-well plates to determine the 50% tissue culture infectious dose per milliliter (TCID
50/mL). Virus titers were calculated using the Reed–Muench method [
30].
2.7. Growth Kinetics Assay
Comparative growth analysis in multiple mammalian cell lines was conducted as described previously [
14]. Briefly, confluent cells were inoculated with AlsPV at an MOI of 0.01 and incubated for 1 h at 37 °C. Cells were washed four times with PBS and cell culture media was added. Infected cells were incubated at 37 °C and aliquots were taken every 24 h for 6 days. Virus titers were determined as above. Virus titers were compared by two-way ANOVA followed by Bonferroni adjustment using GraphPad Prism 5 (LaJolla, CA, USA).
2.8. Immunofluorescence Assay
Confluent Vero cells were infected with virus at an MOI of 0.01 and incubated for 2–3 days at 37 °C. Cells infected with Hendra virus at an MOI of 0.5 were incubated at 37 °C for 24 h under BSL4 conditions. Infected cells were fixed with ice-cold methanol for 15 min, or 30 min for cells infected with Hendra virus, before blocking with 1% BSA at 37 °C for 30 min. Cells were incubated with primary antibody for 1 h at 37 °C and washed four times with PBS-T. Following this, cells were incubated for 1 h at 37 °C with a secondary antibody, either Protein A-Alexa Fluor 488 or anti-rabbit-Alexa Fluor 488, and DAPI. Cells were washed four times with PBS-T before imaging using an EVOS FL Cell Imaging System (Life Technologies, Carlsbad, CA, USA).
2.9. Neutralisation Assay
Paramyxovirus sera were first inactivated by incubating at 56 °C for 35 min. Cell culture media containing 100 TCID
50 of AlsPV or other paramyxoviruses were incubated with two fold dilutions of various paramyxovirus sera or AlsPV ferret sera for 30 min at 37 °C. A suspension of 2 × 10
4 Vero cells was added to each well, and plates were incubated for 5–7 days and then assessed for the presence of CPE. Neutralizing titers were calculated using the Reed–Muench method as described previously as the reciprocal of the highest dilution of serum at which the infectivity of 100 TCID
50 of virus is neutralized in 50% of the wells [
31].
2.10. Australian Flying Fox Serology
Australian pteropid bat sera, collected between 1999 and 2012, were inactivated by treating at 56 °C for 35 min. Sera at a 1:10 dilution were incubated in quadruplicate with 100 TCID50 AlsPV for 45 min before the addition of a suspension of 2 × 104 Vero cells per well. Plates were incubated for 7 days before being assessed for the presence of virus-induced CPE.
2.11. Animal Experiments
All procedures were approved by the CSIRO Australian Animal Health Laboratory Animal Ethics Committee. Study 1, project number 1814, was approved in August 2016. Study 2, project number 1865, was approved in June 2017.
2.11.1. Study 1
Female ferrets (n = 3) were exposed oronasally to 7 × 105 TCID50 AlsV in 1 mL sterile PBS. Adult female BALB/c mice aged between 6–9 months (n = 5) and juvenile (8 week old) female BALB/c mice (n = 5) were exposed intranasally to 2 × 104 TCID50 AlsV in 30 μL sterile PBS whilst under anesthesia (ferrets—0.05 mg/kg medetomidine and 5 mg/kg ketamine; mice 1 mg/kg medetomidine and 75 mg/kg ketamine). Ferrets sourced from the CSIRO Werribee Animal Facility were approximately one year old and had a mean weight of 890 g. Mice were provided by the CSIRO Australian Animal Health Laboratory Small Animal Facility. Virus stock used for animal challenge was diluted 1/10 to reduce the risk of adverse reactions such as laryngospasm during virus challenge, while still maintaining a high enough dose to maximize the likelihood of infection. The challenge doses of inocula were confirmed by back titration.
Animals were monitored for clinical signs of disease for 21 days following challenge. Oral swabs, nasal washes, rectal swabs and EDTA-treated whole blood samples were collected from ferrets on days 3, 5, 7, 10, and 14 days post-infection and again at euthanasia on day 21. Sera were also collected from ferrets on day 7 onward and urine was collected at euthanasia. Weight, rectal temperature, and body temperature measurements were collected from ferrets at each sampling event. Weight and microchip temperature were measured daily for mice.
Tissues collected at euthanasia for assessment by both virus isolation and qRT-PCR were lung, kidney, spleen, brain (olfactory bulb plus 2 mm caudal) and liver from mice; and lung, kidney, spleen, brain (olfactory bulb plus 2 mm caudal), liver and retropharyngeal lymph node from ferrets. Tissues were fixed in a 10% neutral buffered formalin for histology analysis.
2.11.2. Study 2
Female ferrets (n = 12), approximately one year old with a mean weight of 800 g, were exposed to 7 × 105 TCID50 AlsV as for study 1. Three ferrets were euthanized on each of days 3, 5, 7, and 10 post-inoculation, based on random allocation of a time point for euthanasia.
Microchip temperature was recorded daily from all animals following challenge. On the day of euthanasia, oral swabs, nasal washes, rectal swabs, urine, and blood were collected from each ferret, and weight and rectal temperature measurements recorded. Tissues collected for the detection of virus by isolation and qRT-PCR were brain (olfactory bulb plus 2 mm caudal), nasal turbinates, tonsil, trachea, peripheral lung, hilar lung, spleen, kidney, liver, heart, small intestine, large intestine, bronchial lymph node and retropharyngeal lymph node. Tissues were also stored in 10% neutral buffered formalin for histology analysis.
2.11.3. Analysis of Animal Infection Study Samples
RNA was extracted from swab, EDTA-blood and homogenized tissue samples using MagMAX-96 Viral RNA Isolation Kit (Applied Biosystems, Foster City, CA, USA) and analyzed by quantitative RT-PCR. RNA was amplified with AgPath-ID One-Step RT-PCR Reagents (Applied Biosystems) using primers and probe targeting a region in the viral nucleocapsid gene—AlsPV-N287F (5′-AATCCCGAGCTACGTTCAAAACT-3′), AlsPV-N360R (5′-TGGGAGTCACGAGCTCCATT-3′), AlsPV N-311-FAM (5′-FAM-CTGCTATTTTGCCTACGCATTGTGCTGA-TAMRA-3′)—and 18S as an internal control—18S-F (5′-GGCCCTGTAATTGGAATGAGTCCA-3′), 18S-R (5′-GCTGGAATTACCGCGGCT-3′), 18S-VIC (5′-VIC-TGCTGGCACCAGACTTGCCCTC-TAMRA-3′). Reactions were incubated at 45 °C for 10 min and 95 °C for 10 min, and cycled 40 times at 95 °C for 15 s and 60 °C for 45 s on a QuantStudio6 (Applied Biosystems). Copy numbers were calculated using standard curves generated by serially diluting RNA transcribed from control DNA plasmids. To facilitate data interpretation, a copy number of 5 in both qRT-PCR replicates, correlating with a
CT value of 40 (study 1) or 37.4 (study 2), was used as the minimum of detection. Viral N gene copy numbers in tissue samples were standardized to 18S expression (per 10
10 copies of 18S RNA). Viral N gene copy numbers in shedding samples were calculated per milliliter of sample. Results were analyzed using QuantStudio6 software. For virus isolation and titration, 10-fold serial dilutions of samples were made in 96-well plates. A suspension of 2 × 10
4 Vero cells was added to each well and plates were incubated at 37 °C for 7 days before assessing for signs of CPE. Titers were calculated using the Reed–Muench formula [
30].
2.11.4. Histology
Tissues were fixed in 10% neutral buffered formalin, and then trimmed and processed using routine histological methods as previously described [
32]. Sections were assessed for the presence of histopathological lesions and viral antigen following routine hematoxylin and eosin staining and immunohistochemical staining using rabbit antibodies raised against a recombinant AlsPV N protein peptide (Genscript, Piscataway, NJ, USA).
2.12. Antibodies
AlsPV polyclonal antibodies were generated in rabbits by Genscript (USA) using the peptide RQQGRINPRYLLQP from the AlsPV N protein. AlsPV ferret antisera produced in the animal infection trials described in this study were also utilized. Other primary antibodies included rabbit or pig antisera against MenPV (AAHL), rabbit or pig antisera against TioPV (AAHL), TevPV ferret antisera (AAHL), rabbit or horse antisera against HeV (AAHL), PorV rabbit antisera (AAHL) and MapV rabbit antisera (AAHL). Polyclonal anti-PIV5, 21005-2WR (antiserum, guinea pig), NR-3232, polyclonal anti-mumps virus, Enders (antiserum, guinea pig), NR-4019 and polyclonal anti-hPIV2, Greer, (antiserum, guinea pig), NR-3231 were obtained through the NIH Biodefense and Emerging Infections Research Resources Repository, NIAID, NIH.
2.13. Protein Prediction
Membrane topology of AlsPV proteins was predicted using Phobius as described in [
33].
2.14. Sialidase Assay
Confluent Vero cell monolayers were treated with 15 mU of Arthrobacter ureafaciens neuraminidase for 2 h at 37 °C in cell culture media. Untreated and neuraminidase-treated cells were washed twice with PBS and incubated with AlsPV or PIV5 (MOI 2) in duplicate for 1 h. The cells were then washed four times with PBS and incubated for 24 h in cell culture media, before fixing and immunofluorescence staining as above. Fluorescent cells were counted in nine fields of view per well using an ImageXpress Micro XLS Widefield High-Content Analysis System (Molecular Devices, San Jose, CA, USA) and compared to untreated infected cells. The relative number of infected cells was compared by one-way ANOVA followed by Dunnett’s multiple comparison test (compared to untreated cells) using GraphPad Prism 5.
4. Discussion
Analysis of pteropid bat urine collected in northern New South Wales in 2011 has led to the isolation of a novel rubulavirus that we have named Alston virus (AlsPV). Phylogenetic and antigenic analyses indicated that this virus is closely related to PIV5. However, in comparison to the observed variation between AlsPV and PIV5, 26% across the whole genome, isolates of PIV5 are almost identical despite being isolated from a range of host species, geographical locations and over multiple decades [
39]. In fact, variability of only 7.8% is observed between strains of PIV5, with an average pairwise difference of only 2.1% at the nucleotide level [
39]. Given the level of genetic change seen with AlsPV compared to PIV5 (greater than 20% diversity), we propose that AlsPV is a new species of rubulavirus and not a new strain of PIV5.
Despite proposing that AlsPV is a novel species, AlsPV and PIV5 share phenotypic similarities. Experimental intranasal infection of ferrets with PIV5, similar to the infection studies described here, demonstrated variable results ranging from no clinical symptoms to mild cough with minimal lesions in the nasal cavities and upper trachea [
40,
41]. Antigen could only be detected in the trachea with no evidence of virus in the lungs or the brain [
40,
41], although it is not known if the olfactory bulb was specifically investigated. Neurological symptoms have been observed in ferrets experimentally infected with PIV5, but only after intracerebral injection as the route of infection. In addition to these in vivo similarities, PIV5 and AlsPV are similar in that they utilize sialic acid as a receptor for cell entry (49), grow to high titers in multiple mammalian cell lines with minimal cytopathic effect (229, 230) and encode an SH gene (49). The multiple similarities between the two viruses indicate that, despite the lack of clinical disease in ferrets and mice, AlsPV may also have the potential to infect other mammalian species.
The findings presented here suggest that AlsPV causes an upper respiratory tract infection in ferrets that is followed by infection of the olfactory pole of the brain. It is likely that from the nasal turbinates, the virus has access to the olfactory neurons that extend from the olfactory bulb through the cribriform plate and into the olfactory neuroepithelium in the nasal cavity [
42], but further histopathological examination would be required to confirm this hypothesis. This allows direct access for the virus to disseminate through the central nervous system. However, AlsPV was not detected in brain tissue beyond the olfactory bulb. It is likely that the innate immune response had a role in preventing the spread of AlsPV throughout the CNS [
42,
43]. Further investigation into the persistence of AlsPV is required, particularly as reduced cytopathic effect was observed during infection of some mammalian cell lines with AlsPV and viral RNA was detected in the olfactory pole of mouse brains at 21 days post infection.
The timing of detection of virus shedding in respiratory secretions of ferrets suggests that virus replication peaked around 5–7 days post inoculation. Although there was some variability between shedding in the first and second animal infection studies, it may have been due to variation between cohorts of outbred ferrets. The presence of virus in oronasal shedding samples and in the upper respiratory tract suggests that AlsPV replicates in tissues that are relevant to virus transmission, although transmission studies are required to confirm if transmission occurs in ferrets. Further experiments are also required to determine the effect of the route of virus challenge and if oronasal infection with lower doses of AlsPV still results in subsequent infection of the olfactory nerve.
Although AlsPV neutralizing antibodies were found at a low overall prevalence in Australian pteropid bat sera, analysis of individual pteropid species indicated that the
Pteropus poliocephalus flying foxes may be the primary reservoir host. This species was observed in the colony in Alstonville as well as in Geelong where AlsPV was detected by PCR. The high proportion of positive sera samples from grey headed flying foxes, combined with the increasing urban habituation of pteropid bats [
6], suggest that there is risk of exposure and potential transmission of AlsPV to non-pteropid mammalian species.