Background
Respiratory Syncytial Virus (RSV) is the leading cause of lower respiratory tract infection (LRTI) in infants and children world-wide and is increasingly recognised as a cause of serious disease in adults and immune compromised transplant patients [
1,
2]. Over half of all children will be infected with RSV by their first birthday and by the age of 2 nearly all children will have been infected with RSV at least once [
3]. LRTI caused by RSV infection is a major cause of both infant hospitalisation and infant viral induced death [
4]. A number of medical treatments, including use of bronchodilators, palliative care (supportive ventilation, nitric oxide) and use of anti inflammatory agents are available but none of these treatments relieve the viral burden in RSV-infected patients. The only small molecule antiviral therapeutic agent for treating RSV is Virazole (aerosolised ribavirin), which has been shown to be of limited use because of its lengthy administration and questionable efficacy [
5,
6]. Palivizumab (Synagis) is a humanised monoclonal IgG1 antibody specifically directed to the RSV fusion protein which has been used prophylactically to good effect in at-risk infants. However, a therapeutic treatment did not result in significant clinical benefit [
7]. Thus, there is a clear, unmet medical need to develop therapies able to ameliorate RSV disease [
8‐
10].
RSV is a negative-stranded RNA virus belonging to the
Paramyxoviridae family. The negative sense single-strand RSV genome comprises a RNA molecule encoding 11 proteins. Upon host cell infection positive-sense viral mRNAs are synthesised by the viral RNA polymerase, these mRNAs make use of host-cell machinery to synthesise viral proteins. Genome replication occurs via the production of a positive sense replicative intermediate (RI) RNA strand by the same viral RNA polymerase; this RI RNA is used as a template for the synthesis of more negative sense genome [
11‐
13].
The use of
in vivo models with good clinical translation is vital in the search for new treatments for disease A number of different animal models have been used to study RSV infection and replication and to evaluate potential therapies, including primates, bovines and rodents [
14]. The majority of
in vivo studies have been conducted using either the BALB/c mouse [
15] or the cotton rat (
Sigmodon hispidus) [
16] models. The cotton rat is moderately permissive to human respiratory viral infection and RSV is able to replicate and produce viral progeny in the lungs [
17]. The BALB/c mouse is also susceptible to RSV infection [
18] and, though less permissive than the cotton rat [
19], constitutes a more practical model due to the availability of a larger number of immunological and molecular reagents as well as the availability of transgenic animals. Like the cotton rat, the mouse requires inoculation with a high dose (usually 10
6 PFU) to achieve viral replication. The actual amount of viral replication occurring following infection with such a supra-physiological dose of RSV has never been accurately determined.
We therefore developed a strand-specific real-time quantitative polymerase chain reaction (QPCR) method to monitor the kinetics of RSV RNA replication in the mouse lung. BALB/c mice were infected with RSV A2 and viral RNA in mouse lungs were monitored over an extended time course. Levels of infectious virus in lungs were also measured. Taken together, results from these 2 assays showed that RSV RNA synthesis and viral replication was severely limited in the mouse. Treatment with a prophylactic antibody (palivizumab) did not affect viral RNA replication and persistence, but did impair the production of infectious progeny virus, indicating that abortive replication [
16] occurs in the mouse. By contrast, positive sense viral RNA and infectious virus production were both disrupted by ribavirin. Further
in vitro studies in human and mouse cells demonstrated that although both cell types were equally susceptible to infection; viral RNA synthesis was delayed and impaired in mouse cells. This finding suggests that a species-specific host-virus interaction inhibits the capacity for RSV replication in the mouse.
Methods
Animals
Female BALB/c mice (6-8 weeks old), specific pathogen free, were purchased from Charles River Laboratories and housed in an animal care facility in ventilated isolation cubicles. Water and chow were provided ad libitum. Mice were allowed to acclimate to the new environment for 1-2 weeks and housed in groups according to experimental setup. All experiments with animals were carried out in compliance with UK legislation and subject to local ethical review.
Virus, cells and viral assays
RSV-A2 was obtained from Advanced Biotechnologies Inc. Stocks were produced by infecting Hep-2 cells at a multiplicity of infection (MOI) of 0.1 focus forming units (FFU) per cell. Following 4-5 days incubation, infected cells were harvested and snap frozen in dry ice and methanol and stored at -80°C. Viral titres were determined by a HEp2 based immunofluorescence assay and expressed as FFU/ml [
20]. UV-inactivated RSV (UVRSV) was generated by exposing RSV A2 to UV radiation at 254 nm for 20 minutes using a Stratalinker (Stratagene). Loss of infectivity of UVRSV was confirmed by infecting Hep2 cells (MOI ranging from 0.1-1 FFU/cell). For animal studies, viral titres were expressed as geometric means +/- standard errors of means (SEM) for all animals in a group.
A549 cells (human lung carcinoma) and KLN205 cells (DBA/2 mouse lung squamous cell carcinoma) were purchased from ATCC and maintained in DMEM or EMEM respectively, each supplemented with 100 IU/ml of penicillin, 100 μg/ml streptomycin, 2 mM L-glutamine, and 10% foetal calf serum (FCS).
Drugs
Ribavirin was obtained from Sigma-Aldrich and Palivizumab (Synagis, MedImmune) was obtained from Idis Ltd.
RSV infection in vivo
For preliminary RSV replication and dynamics studies, mice were inoculated once intranasally (i.n.) with 50 μl of either RSV A2 (1 × 10
6 FFU per animal) or an equivalent concentration of UVRSV. One group of control mice was left untreated. Animals were sacrificed at 1, 5, 8, 17, 24, 48 and 72 hours and 7, 10, 37 and 59 days after infection (3 mice per time point). The lungs were removed from the thorax, dissected into two and each weighed. One lung was placed into RNAlater (Ambion) for subsequent RNA extraction and Taqman analyses. The second lung was processed by hand-held homogeniser (Omni) in 1 ml MEM (Invitrogen). Homogenates were centrifuged, clarified viral supernatant diluted 1:3 in MEM and 50 μl used in triplicate in immunofluorescence assay [
20].
For experiments conducted to investigate inhibition of RSV replication, one group of animals were administered a single intramuscular injection of palivizumab (5 mg/kg of body weight) 24 hours prior to infection with RSV. A second group were administered ribavirin (100 mg/kg of body weight) intraperitoneally one hour prior to RSV challenge. These groups, plus a further untreated group, were inoculated intra nasally with 75 μL RSV A2 (2.6 × 106 FFU/mouse). Ribavirin treatment was re-administered 5 hours post virus inoculation and twice daily dosing of this compound continued for a further day. Ribavirin treatment was not administered on day 2. Dosing continued on day 3 at 50 mg/kg twice daily until day 6 post virus infection. Animals were sacrificed at 1, 8, 24, 48 and 72 hours and 5, 7 and 10 days post infection (6 animals per group per time point). Lungs were harvested for viral titrations and RNA extraction.
In-vitro transcript standard production
A region of the RSV A2 nucleocapsid domain was isolated using a nested primer approach. RSV A2 viral RNA was prepared from crude preparation using the QIAamp viral RNA minikit (Qiagen). RNA was reverse transcribed using the High Capacity cDNA reverse-transcription kit (Applied Biosystems) with random primers. PCR was conducted using
Pwo Superyield polymerase (Roche) with external primers (Table
1) at an annealing temperature of 60°C for 35 cycles followed by nested primer (Table
1) PCR using cycling conditions as described above. Gel-purified PCR product was restriction-cloned into pGEM-4Z vector (Promega), grown in Oneshot TOP10 chemically competent
E. coli (Invitrogen) and plasmid purified by QIAprep Spin Miniprep Kit (Qiagen). Clones were sequence-checked at Lark Technologies, UK. The insert plus bacterial promoter vector sequences of verified clones was isolated by PCR using
Pwo Superyield polymerase with M13 forward (-20) and reverse primers at an annealing temperature of 55°C for 35 cycles. Positive and negative sense in vitro transcripts were synthesised by Sp6 and T7 RNA polymerase (Promega) respectively, these products were treated with Turbo DNase (Ambion) and purified by 3 M sodium acetate (pH 5.5) precipitation. Stocks of 10
8 absolute copies per μl were prepared and stored at -80°C.
Table 1
Primer sequences used for RNA standard generation, cDNA synthesis and QPCR.
In vitro standard external positive sense | TCCAGCAAATACACCATCCA |
In vitro standard external negative sense | CTGCTTCACCACCCAATTTT |
In vitro standard nested positive sense | ATAGAATTC GGTATGTTATATGCGATGTCTAGGT1 |
In vitro standard nested positive sense | ATAGGATCC TGCTAAGACTCCCCACCGTAA2 |
Positive sense RNA-specific cDNA synthesis | CGGTCATGGTGGCGAATAA TCCTGCAAAAATCCCTTCAACT3 |
Negative sense RNA-specific cDNA synthesis | CGGTCATGGTGGCGAATAA ACTTTATAGATGTTTTTGTTCA3 |
Positive sense-specific QPCR primer | CCCCACTTTATAGATGTTTTTGTTCA |
Negative sense-specific QPCR primer | TCCTGCAAAAATCCCTTCAACT |
QPCR tag primer | CGGTCATGGTGGCGAATAA |
Probe | FAM-TTGGTATAGCACAATCTTCTACCAGAGGTGGC-TAMRA |
Strand-specific real time QPCR
RNA was prepared from mouse lungs using an RNeasy kit (Qiagen) following manufacturer's instructions. First strand cDNA was synthesised from RNA using Reverse Transcription Reagents (Applied Biosystems) with gene specific primers targeted to the positive or negative sense RSV A2 nucleocapsid region RNA (Table
1). Primers contain a tag sequence recognised by a tag-specific primer in QPCR reactions; this reduces the detection of non-specific, self-primed cDNAs [
21]. Reactions (10 μl) comprised 1 × reaction buffer, 5.5 mM MgCl
2, 0.5 mM dNTP mix, 2.5 μM strand-specific primers, 4 U RNase inhibitor and 12.5 U reverse transcriptase with 4 μl total RNA preparation in water. Reactions were performed at 50°C for 40 mins followed by 95°C for 5 mins. Positive strand detection by QPCR was performed using TaqMan
® Universal PCR mastermix (Applied Biosystems) with positive sense RNA specific primer, 800 nM tag-specific primer and 100 nM probe (Table
1). Reactions were performed using an Applied Biosystems 7900 HT. Samples were held at 50°C for 2 mins followed by 95°C for 10 mins and then 40 cycles of 95°C for 15 secs and 60°C for 1 min. Negative sense strand detection was performed as described for the positive sense RNA reaction but substituting the positive sense RNA specific primer for a negative sense RNA primer. Positive and negative sense RNA transcript standard ranges (10-10
7 absolute copies/μl) were processed alongside samples. The limits of detection for this assay were defined as values measured outside the range of the standard curves. RSV copy number per μl of total mouse lung RNA were normalised to beta-actin detected using commercially available TaqMan
® VIC/MGB primer-limited endogenous control (Applied Biosystems) with random-primed 1
st strand cDNA synthesised using the High Capacity cDNA reverse-transcription kit (Applied Biosystems). Absolute values of normalised RSV copy number were subsequently divided by the weight of the lung tissue from which RNA was extracted and expressed as normalised copy number/g lung wt.
RSV infection in vitro
To investigate whether RSV RNA synthesis occurs effectively in a mouse cell line compared to a human cell line in vitro, human lung carcinoma cells (A549) and mouse lung epithelial squamous cells (KLN205) were plated at a density of 1 × 104 cells per well in 96 well plates and infected with RSV A2 to yield various multiplicities of infection (MOIs) ranging from 1 × 10-3 to 1. Media containing 10% FCS was replaced with fresh media containing 2% FCS after 24 hours. Cells were lysed with RLT buffer (Qiagen) at 1, 8, 24, 48 and 72 hours and after 5 (A549) or 6 (KLN205), 7 and 10 days. Total RNA was prepared using the RNeasy 96 kit (Qiagen). RSV strand-specific QPCR was performed as described above. RSV copy number per μl of total RNA were not normalised to beta-actin but rather analysed separately due to variable rates of cell death observed throughout the experiment and expressed as RSV copy number.
Statistics
For QPCR analyses the ratio of positive to negative copy number is analysed on the logarithmic scale. Treatments are compared to untreated RSV infected controls at each time point by two-sample t-test incorporating Satterthwaite's adjustment to the degrees to freedom. To allow for testing of multiple time points within a treatment a Bonferroni adjustment was made to achieve an approximate 5% significance level within that treatment. Infectious virus assay data were analysed by 1 way analysis of variance (ANOVA) for significant differences (p = < 0.05) between treated groups and untreated RSV infected controls at each time point.
Discussion
We have developed a strand-specific QPCR method to measure RSV in vitro and in vivo. This method distinguishes between negative sense viral RNA (genome) and positive sense RNA (replicative intermediate and nucleocapsid mRNA). Using this method, we provide a detailed insight into RSV RNA production in infected BALB/c mouse lung. To our knowledge, this is the first time that a strand specific method has been applied to profile RSV RNA dynamics in the BALB/c mouse over such a detailed time course.
Early viral RNA synthesis in mouse lungs is characterised by absolute measures of positive and negative sense RNA being equivalent at infection, followed by a 1-2 logs relative increase in positive strand RNA by day 3 post infection. This disparity between RNA strands decreases again from day 7. It should be noted that this window of maximum disparity between the positive and negative strand copy numbers at day 3 coincides with the highest level of infectious progeny virus detected from mouse lungs following infection. It is known that paramyxovirus replicative intermediate RNA represent 10-40% of the genome [
16], therefore the majority of positive strand RNA synthesis seen here is accounted for by nucleocapsid mRNA production.
That RSV genome and positive strand RNA can be detected in mouse lungs up to at least 59 days post-infection has been reported both here and elsewhere [
15,
23]. It therefore appears that mice are unable to fully clear the virus following infection. The fact that UV killed RSV was not detected by QPCR past day 7 supports this view of viral persistence. RSV persistence in the lungs has been reported from humans with chronic obstructive pulmonary disease (COPD) [
24], although in another study, RSV infections in COPD were attributed to acute infection rather than low-level persistence [
25]. The significance of persistent low levels of RSV in this and other conditions is unclear at present and further studies are required to elucidate the scope and impact of this phenomenon [
26]. However, it is possible that low levels of persistent virus exist between RSV seasons and it is apparent that RSV persistence and strategies for complete viral clearance may be studied in rodent models.
Viral RNA replication has been studied by strand-discriminate QPCR previously in the cotton rat [
16]. Viral genome levels increased by approximately 2 logs from 6 hours post infection to a peak measured on day 4 whereas our studies indicate that in the mouse lung total genomic RNA did not increase in this time frame. Indeed, in the mouse model we observed that viral genome load either decreased after 24 hours or (if a higher inoculum was applied), was maintained for a period of time before decreasing after day 5. These data suggest that RSV has a greater replicative capacity in the cotton rat model compared to the mouse. However until a direct head to head comparison is made between the two species, this cannot be concluded.
Ribavirin has been used extensively as an antiviral therapeutic. Its exact mode of action is poorly defined although several mechanisms have been proposed [
27]. Here, as expected, ribavirin treatment had a marked effect on RSV intracellular RNA dynamics as evidenced by the reduction in positive sense RNA in mouse lungs. However, there was little difference seen in the time-course profiles of total genomic RNA in ribavirin treated and untreated RSV infected mice. This suggests that the amount of new genome synthesised following infection is only a small fraction of that dosed initially and that measuring positive sense RNA specifically is vital to the study of the intracellular viral processes in mouse lung following supra-physiologic dosing.
Prophylactic treatment of RSV-infected mice with the neutralising antibody palivizumab resulted in a reduction in infectious progeny virus detected in the lung, although a reduction in positive sense strand RNA was not observed. These findings agree with those previously observed in the cotton rat, where a lack of detectable progeny virus occurred despite intracellular replication taking place. This phenomenon was termed abortive replication [
16]. The authors speculated that abortive replication could occur due to the blocking of production and release of large amounts of progeny virus despite infection occurring in the presence of high titres of neutralising antibody. Our data support this hypothesis. We conclude that the evaluation of antibody-mediated viral therapies in the mouse model may be confounded by the high viral titres required for effective infection.
To investigate whether the restricted replication pattern seen in the mouse is purely an
in vivo phenomenon, we infected lung epithelial carcinoma cells from human (A549) and mouse (KLN205) with RSV and studied viral replication by strand-specific QPCR. One hour post infection, the input viral RNA levels were very similar in both human and mouse cells, irrespective of MOI or cell type, indicating that the mouse and human cells had been exposed to equivalent amounts of viral RNA. However, a clear increase in either viral RNA strand only occurred in mouse cells when they were infected with a high MOI of 0.1 or 1. This situation mirrors that which occurs in the mouse
in vivo model in that an extremely high viral titre is required for replication [
14]. Moreover, the increase in positive strand viral RNA was considerably delayed, occurring after a lag time of 3 days in culture suggesting that the virus has undergone a period of adaptation. Overall, RSV RNA synthesis in human A549 cells was at least 3 orders of magnitude more efficient than that observed in mouse cells, illustrating that RSV cannot replicate efficiently in mouse KLN205 cells. This data suggests that some host-specific block to viral replication exists, though a wider range of human and mouse cell lines require testing to confirm this.
It is unclear why the murine cells did not facilitate RSV RNA synthesis to the same extent as seen in human cells. It may be that RNA replication in KLN205 cells is inhibited either by the presence or absence of one or more host factors required for the viral life cycle. For example, it is known that RSV can modulate host cell anti-viral responses, such as the degradation of STAT2 by NS1 [
28], which inhibits the interferon response. Poor replication of RSV in mouse embryo cells has been described previously [
29]. This was attributed to the mouse interferon response as treatment of infected cells with anti-mouse interferon improved virus yields. Perhaps RSV is not able to modulate the mouse interferon response to the same extent as human interferon. Alternatively, it is also known that RSV requires host proteins to replicate efficiently. Phosphorylation of the RSV P protein by casein 2 is required for transcription elongation activity of the viral polymerase
in-vitro[
30]. It is plausible that species-specific differences in host factors may impair the ability of RSV to replicate efficiently in mouse cells, as is exemplified with HIV and APOBEC3G [
31].
In conclusion, we have demonstrated and quantified the abortive and restricted nature of RSV RNA synthesis and replication in mouse using a highly sensitive and specific QPCR method. We have gone on to provide evidence that the impaired replication may be due to a murine host-virus interaction. We suggest a number of candidates and work is ongoing to identify these interactions.
Acknowledgements
The authors thank Julien Browne, Frances Burden, Bhavika Desai, Tansi Khodai, Susanne Lang, Hannah Perkins and Joanne Strawbridge for practical support. We would also like to thank Chloe Brown, Lisa-Marie Burrows and Lindsey Cousens in Pfizer CM. The statistical support of Katrina Gore and Richard Lyons is also gratefully acknowledged.
Competing interests
All authors are or were employed in a full-time capacity by Pfizer Research and Development.
Authors' contributions
RB carried out the molecular and cellular studies and drafted the manuscript. DR carried out the in vivo and cellular assays and analysis and interpretation of data, EJM, MW and CL participated in the design of the study and analysis and interpretation of data. HB conceived of the study, participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.