Henipaviruses are negative stranded viruses that belong to the Paramyxoviridae
family and are hosted by several species of pteropid fruit bats [1
]. Two henipaviruses, Hendra virus (HeV) and Nipah virus (NiV), cause respiratory disease and encephalopathies with high case-fatality rates in humans and are classified as biosafety level 4 (BSL-4) pathogens [3
]. HeV was identified and isolated in the 1990s after the severe disease in horses and equine workers, and subsequent work identified pteropus bats (Pteropus alecto
and P. conspicillatus
) as reservoir hosts of the virus [5
]. NiV was first identified in Malaysia during an outbreak of severe disease amongst abattoir workers that had contracted the virus from swine. Field studies subsequently identified that other species of pteropus bats were reservoir hosts of NiV [6
]. Periodic outbreaks of henipavirus disease continue to occur in Australia and Southeast Asia. Since the discovery of these viruses, henipavirus sequences have been detected in many pteropid fruit bat species in Australia, Asia, and Africa, posing a significant risk to humans and livestock.
During the processing of pooled urine samples collected from pteropus bats, a novel henipavirus, Cedar virus (CedPV), was isolated in a cell culture [7
]. Sequencing the CedPV genome indicated that its phosphoprotein gene (P) lacked the RNA editing site commonly found in other paramyxovirus genomes, including HeV and NiV, that results in the expression of the V and W accessory proteins [7
]. The V and W proteins have been shown to have IFN-antagonism capabilities in cells infected with NiV [8
]. CedPV infection does not result in disease in guinea pigs or ferrets, both of which are commonly used in henipavirus pathogenesis studies [7
Pathogenic henipaviruses infect many cell types, and the targeting and dysregulation of vascular endothelial cells is a major cause of pathogenesis [9
]. These viruses bind ephrin B2 and ephrin B3 expressed on various subsets of endothelial cells, and the spatial expression of these receptors is associated with viral tropism [11
]. Ephrin B2 is expressed on cells throughout the body; however, ephrin B3 is primarily expressed on cells in the CNS, and this expression pattern is likely responsible for the CNS tropism of NiV and HeV. CedPV binds to ephrin B2 but not ephrin B3, which may account, in part, for its lack of pathogenicity in rodent animal models [7
NiV and HeV infections cause a fatal disease in Syrian golden hamsters (Mesocricetus auratus
) with pathogenesis involving the respiratory tract and nervous system similar to that of the human disease [14
]. We, thus, sought to determine if Syrian hamsters are susceptible to CedPV, whether there are differences in viral replication and immune activation between NiV and CedPV in vitro, and whether these were attributable to CedPV’s inability to cause disease. We compared the immune responses to infection with NiV and CedPV in a cell line deficient in IFN responses and in primary endothelial cells. Both human and hamster primary endothelial cells, the targets of henipavirus infection, generated IFN responses to CedPV that were not observed upon NiV infection. The infection of hamsters by NiV results in severe disease; however, CedPV infection resulted in an immune response that cleared the virus without disease. Together, these data suggest that CedPV activates innate immune responses and is unlikely to cause disease in an immunocompetent host, which is likely due, in part, to the lack of V and W accessory proteins.
2. Materials and Methods
2.1. Ethics Statement
The Institutional Animal Care and Use Committee of the Rocky Mountain Laboratories approved all animal experiments, which were performed following the guidelines of the Association for Assessment and Accreditation of Laboratory Animal Care, International (AAALAC) by certified staff in an AAALAC-approved facility (protocol 2013-059E). All infectious virus work was conducted in the biosafety level 4 (BSL-4) laboratory at Rocky Mountain Laboratories, NIAID (Hamilton, MT, USA) following established BSL-4 standard operating procedures that were approved by the Institutional Biosafety Committee. The isolation of primary microvascular endothelial cells was approved by the Colorado State University Institutional Animal Care and Use Committee (14–4976A, 25/04/2014).
2.2. Viruses and Cells
The Bangladeshi strain of the Nipah virus (NIV) was obtained from the Viral Special Pathogens Branch of the CDC, was originally isolated from a fatal human infection in Bangladesh in 2004, and was passaged in Vero E6 cells a total of three times [16
]. Cedar virus (CedPV) was obtained from CSIRO (Geelong, Australia) and was isolated from bat urine inoculated onto primary bat kidney cells, followed by passaging on Vero E6 cells (ATCC) [7
]. Both viruses were propagated on Vero E6 cells at RML in Dulbecco’s minimal essential medium (DMEM) (Sigma) supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 50 IU/mL penicillin, and 50 µg/mL streptomycin (Life Technologies, Carlsbad, CA, USA).
BHK-21 cells (CCL-10, ATCC, Manassas VA, USA) were propagated in 24-well tissue culture plates and inoculated with CedPV or NiV (or mock) with a MOI of 1.0 or 0.1 for 1 h, washed twice in DMEM, and incubated at 37 °C, 5% CO2 for the indicated amount of time.
Human lung blood microvascular endothelial cells (LBMVEC) (CC-2527, Lonza, Walkersville MD, USA) were maintained in an EGM-2 medium (Lonza). The cells were seeded in 24-well plates one day prior to inoculation with CedPV or NiV with a MOI of 0.1. At the indicated time points, supernatants were collected for virus quantitation and the cells were lysed in buffer RLT (Qiagen) to quantitate viral or cellular RNA.
2.3. Hamster Infections
Five- to six-week-old female Syrian golden hamsters (Harlan, Indianapolis, IN, USA) were inoculated intraperitoneally (IP) (a total of 500 µL) or intranasally (IN) (a total of 100µL) with 105 TCID50 of CedPV that was diluted in sterile DMEM. CedPV was passaged twice on P. alecto kidney cells upon original isolation from bats and passaged twice on Vero E6 cells to generate a virus stock. The control animals received DMEM only. At the indicated time points, groups of 4 hamsters were euthanized and the lung tissue was collected for virus quantitation and histology, and sera was collected for the measurement of antibody titers by a neutralization assay.
2.4. Virus Quantification
RNA was extracted from homogenized tissue samples using the RNeasy Mini Kit (Qiagen, Carlsbad, CA, USA) following manufacturer’s instructions and according to our established protocols. Viral RNA was quantitated using a one-step real-time RT-PCR targeting NP genes with a Rotor-Gene probe kit (Qiagen) (primer and probe sequences available on request). Standard dilutions of the RNA extracted from the titrated virus stocks were assayed in parallel to calculate the FFU equivalents of viral RNA.
The tissue culture infectious dose 50% (TCID50) method was used to quantitate CedPV and NiV in the supernatants of the infected cell cultures. Monolayers of the Vero E6 cells were grown in 96-well plates, and 100 µL of the serial 10-fold diluted samples in MEM containing 2% FBS were added to the wells in triplicate. The cells were then incubated for 5 days at 37 °C with 5% CO2 and then scored for CPE, and the Spearman–Karber method was used to calculate the TCID50.
2.5. Virus Neutralization Assay
The neutralizing antibody titers were measured by a neutralizing tissue culture infectious dose 50% assay (NTCID50). For this, 100 TCID50 of CedPV was incubated with serial dilutions of hamster sera for 1 h at 37 °C. This was used to inoculate monolayers of the Vero E6 cells in 96-well plates for 1 h. After five days of incubation, the CPE was examined and the NTCID50 was calculated using the Spearman–Karber method.
2.6. Generation of Primary Hamster Pulmonary Microvascular Endothelial Cells
To obtain pulmonary endothelial cells from Syrian hamsters, the protocol from reference [17
] was modified. Briefly, 5-day-old hamster littermates were anesthetized with isoflurane to effect before euthanasia by thoracotomy. Immediately after euthanasia, the lungs were aseptically collected and placed in ice-cold, serum-free DMEM. The lungs were then transferred to sterile Petri dishes, and the excess DMEM was aspirated. Tissues were minced by repeated cutting with sterile scissors approximately 100 times before enzymatic digestion with a collagenase/elastase (C1639, E7885, Sigma, St. Louis, MO USA) cocktail at 37 °C on an orbital rocker for 45 min. The digested cell suspension was passed through a 20 g needle 12 times to generate a single cell suspension before an isolation medium (20% FBS, 1% pen/strep in DMEM) was added to neutralize the enzymatic digestion. After tissue digestion, the cells were pelleted by centrifuging at 400× g
for 5 min and then washed in PBS containing 0.1% BSA. The cells were resuspended in 100 µL of 0.1% BSA-PBS (Ca2+
-free). A magnetic separation was used to select for endothelial cells. Fifty microliters of magnetic beads conjugated to Protein G (G7471, Promega, Madison, WI, USA) were washed 4 times in 0.1% BSA-PBS before resuspension in 0.1% BSA-PBS. The anti-CD31 (PECAM1) antibody (LS-C150165, LSBio, Seattle, WA, USA) was conjugated to the magnetic beads by mixing 50 µL of the antibody with 100 µL of the magnetic bead slurry for 2 h at room temperature on an orbital mixer. The antibody-conjugated beads were then washed twice in 0.1% BSA-PBS before being resuspended in 500 µL of 0.1% BSA-PBS. To sort the cells, 250 µL of the antibody-conjugated beads were added to 1 mL of the resuspended cells and tubes were tumbled on an orbital mixer for 15 min at RT. The samples were then placed on a magnet, and the cells not bound to the magnetic beads were aspirated and discarded. The cells bound to beads were washed 4 times in 0.1% BSA-PBS before being resuspended in 1 mL of the VascuLife EnGS Mv endothelial cell culture media (LL-0004, LifeLine Cell Technology, Carlsbad, CA USA). The cells were plated in 35-mm cell culture dishes that had been pre-coated in gelatin (G1393, Sigma, St. Louis, MO, USA), and an additional 1 mL of media was added. Half of the medium (1 mL) was changed every other day until the cells had grown to 80–90% confluency, at which time the cells were expanded and used in experiments.
2.7. Real-Time PCR of Host Response Genes
SYBR Green real-time PCR was performed using primers for a subset of innate response genes (Supplemental Table S1
). Briefly, the primers were designed using software developed by us [18
] from the Syrian hamster RefSeq RNA database. Primary hamster endothelial cells were infected with 0.1 MOI of CedPV or NiV for 1 h followed by medium replacement (2% FBS-DMEM). The total RNA was extracted from endothelial cells (RNAEasy, Qiagen) at 24 and 48 h postinfection and reverse transcribed (QuantiTech RT, Qiagen) according to manufacturer’s instructions, followed by qPCR (QuantiTech SYBR Green Master Mix, Qiagen) with a 96-well LightCycler (Roche). The cycling parameters were a 15 min activation step followed by 40 cycles of 15 s at 94 °C, 30 s at 56 °C, and 30 s at 72 °C. Melt curves were generated post-amplification to ensure only single-peak products were generated. The relative template abundances were determined using the ΔΔCt method normalized on GAPDH.
Previous work showed that, unlike NiV, CedPV fails to cause disease in ferrets or guinea pigs [7
]. Because of the number of immunological reagents and the available annotated genome for Syrian hamsters, which are model organisms that recapitulate many aspects of human disease caused by NiV and HeV, we challenged hamsters with CedPV and determined that they were susceptible to infection but do not develop disease. The inoculation of hamsters indicates the intranasal route is more efficient at inducing infection than is the intraperitoneal route (Figure 1
). Despite the lower frequency of infection via the intraperitoneal route, those hamsters that became infected developed neutralizing antibody responses that paralleled those infected intranasally (Figure 2
). The replication of CedPV in the lungs of intranasally inoculated hamsters at early time points (2–4 days postinfection) was approximately 102
lower than what we observed for NiV at a similar time point (5 days) [20
]. Comparisons after this time point are not possible because hamsters succumb to infection between 5–7 days postinfection. To elucidate the innate immune differences between the pathogenic NiV and apathogenic CedPV, we performed cell culture experiments using hamster and human cells. The lack of pathogenicity caused by CedPV, in contrast to NiV/HeV, correlated with the induction of CedPV-induced innate immune responses.
Nearly all paramyxoviruses, including NiV and HeV, possess a phosphoprotein (P) gene that is unusual in that it has an RNA editing site resulting in frameshifts that lead to alternative reading frames that encode three other polypeptides: C, V, and W [21
]. However, CedPV is devoid of this RNA editing site and is incapable of producing V or W proteins. The loss of this site appears to be natural and is not a result of passaging in the cell culture because there are no coding regions in the P gene that are similar to V or W of HeV or NiV [13
]. The V and W proteins interfere with the type I IFN response by disrupting the STAT1 and STAT2 localization to the nucleus [8
]. In BHK-21 cells, a fibroblast line incapable of secreting type I IFN [26
], both CedPV and NiV replicated with similar kinetics and to similar titers (Figure 3
). However, in type I IFN-competent hamster primary cells, CedPV replication was reduced by more than 4 logs compared to NiV and to CedPV in BHK-21 cells (Figure 4
), suggesting that CedPV is unable to suppress the type I IFN response, resulting in an antiviral state that tempers replication. This was recapitulated in human primary pulmonary endothelial cells.
We hypothesized that the coinfection of hamster endothelial cells (Figure 5
) would rescue CedPV replication to levels found in BHK-21 fibroblast cells (Figure 3
) because NiV suppresses the type I IFN response. However, CedPV replication remained at the same level as with virus alone in these cells (compare Figure 4
A to Figure 5
A). There are two possible explanations for the failure of NiV coinfection to support a greater CedPV replication in primary endothelial cells. First, there were insufficient numbers of dually infected cells with a NiV-mediated block of the innate immune mechanism in CedPV-infected cells. This outcome may be due to insufficient numbers of coinfecting particles or may result from superinfection exclusion [27
]. Such an exclusion would effectively prohibit the dual infection of individual cells and prevent the protection of CedPV by NiV. A second possible mechanism is cell type-dependent differences in RNA replication kinetics. The CedPV and NiV polymerase complexes may be viral RNA-specific and rate limiting in endothelial cells. Paramyxovirus polymerase complexes are composed of the polymerase (RdRp), nucleoprotein (N), and the phosphoprotein (P). The nucleoproteins of the CedPV and NiV share only a 66% identity, and the RdRp share only a 57% identity but CedPV’s RdRp is 256 amino acids larger [7
]. The P proteins share only a 34% identity with the great majority of diversity between residues 200–400. Thus, there are sufficient differences between the polymerase complexes to suggest that their function would be restricted from the replication of a coinfecting virus in trans
. This explanation also requires that a restricted function of the CedPV RdRp occurs specifically in primary endothelial cells. However, in BHK-21 cells, a maximal CedPV polymerase activity leads to a high titer replication.
To further scrutinize the impact on the type I IFN response in hamster endothelial cells, we examined the gene expression profiles of several interferon response genes by qPCR. NiV effectively suppressed the expression of these genes, whereas CedPV did not (Figure 6
). Together, these data suggest that an inability of CedPV to circumvent the type I IFN response may contribute to its reduced replication and limits its ability to cause disease in hamsters.
Cellular attachment by HeV and NiV is restricted to ephrin B2 and ephrin B3 [7
]. It is thought that ephrin B3 binding is a prerequisite for encephalitis in humans and other mammals because it is expressed by cells of the brain stem, whereas ephrin B2 is principally found on other cells, including endothelial cells. Because CedPV selectively binds ephrin B2, it is likely unable to infect cells of the brain stem. However, it is unclear whether the combination of type I IFN induction, lower replication, and lack of ephrin B3 tropism leads to an apathogenic transient CedPV infection. Future studies using introduced mutations of the recently described reverse-engineered infectious clone of CedPV [29
] will be essential for addressing these questions.
The loss of the V and W proteins in CedPV also raises important biological questions about their roles in the reservoir bat species. The clear evidence that CedPV circulates amongst pteropid bats in Australia [30
] implies V and W are dispensable for persistence in bats, some of which have constitutively activated interferon pathways. The loss of V and W proteins should be detrimental to CedPV replication. Its persistence implies the existence of other viral and host factors that contribute to CedPV ecology.