- Email: [email protected]
Mechanisms of temperature sensitivity of attenuated Urabe mumps virus
Accepted Manuscript Title: Mechanisms of temperature sensitivity of attenuated Urabe mumps virus. Author: Stephanie C.Burke Schinkel Steven Rubin Kathryn E. Wright PII: DOI: Reference:
S0168-1702(16)30551-2 http://dx.doi.org/doi:10.1016/j.virusres.2016.10.003 VIRUS 96971
To appear in:
Virus Research
Received date: Revised date: Accepted date:
30-8-2016 4-10-2016 5-10-2016
Please cite this article as: Schinkel, Stephanie C.Burke, Rubin, Steven, Wright, Kathryn E., Mechanisms of temperature sensitivity of attenuated Urabe mumps virus.Virus Research http://dx.doi.org/10.1016/j.virusres.2016.10.003 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Mechanisms of temperature sensitivity of attenuated Urabe mumps virus. Stephanie C. Burke Schinkela , Steven Rubinb and Kathryn E. Wrighta* a
Department of Biochemistry, Microbiology and Immunology, University of Ottawa 451 Smyth Road, Ottawa, Canada K1H 8M5 b Food and Drug Administration (FDA), Center for Biologics Evaluation and Research (CBER), OVRR, DVP, 10903 New Hampshire Avenue, Bldg 52/72, Silver Spring, MD, 20993 [email protected] [email protected] [email protected] *corresponding author
Highlights
Attenuation of viruses is often linked to temperature sensitivity A temperature sensitive isolate of Urabe mumps virus was shown to be attenuated Genetic substitutions unique to this virus were identified including changes in the virus polymerase Functioning of the viral RNA polymerase was reduced at the non-permissive temperature
Abstract Temperature sensitivity is a phenotype often associated with attenuation of viruses. Previously, we purified several mumps variants from an incompletely attenuated Urabe strain live attenuated vaccine. Here we characterize one isolate that is sensitive to growth at high temperature. This virus was attenuated in a small animal model of mumps virulence, and we identified unique coding substitutions in the hemagglutinin-neuraminidase (HN), the viral polymerase (L) gene, and a non-coding substitution close to the anti-genome promoter sequences. At the non-permissive temperature, transcription of viral mRNAs and production of the replication intermediate were reduced compared to events at the permissive temperature and to a non-ts virulent Urabe virus. As well, synthesis of viral proteins was also reduced at the higher temperature. While the actual sequence substitutions in the ts virus were unique, the pattern of substitutions in HN, L and genome end sequences is similar to another attenuated Urabe virus previously described by us.
Keywords Mumps virus, attenuation, temperature sensitivity, transcription, replication
The genetic basis of virus attenuation (att) is complex in that each attenuated virus may have unique mutations, but attenuating mutations often have similar functional implications. We have shown that the AM9 Urabe (Ur) mumps vaccine, which was incompletely attenuated, contained several genetic variants of the Ur virus (Brown et al., 1996). Plaque purified isolates from the vaccine displayed differences in growth and fusion properties in tissue culture (Wright et al., 2000), displayed differences in attenuation in the rat neurovirulence model (Shah et al., 2009), and at least two viruses were temperature sensitive (ts). We had reasoned that the ts variants were likely amongst the most attenuated variants within the vaccine (Wright et al., 2000). Ts mutants of viruses commonly arise during attenuation processes such as serial passage of viruses in non-host cells (Saika et al., 2006, Sakata and Nakayama, 2011), codon-pair deoptimization (Le Nouën et al., 2014) and cold adaptation (Karron et al., 1997, Hall et al., 1992), and the ts phenotype is linked to attenuation in vivo (Crookshanks and Belshe, 1984). It is not clear how mutations attributing the ts phenotype also affect attenuation. For some viruses, particularly those that are also cold adapted (ca), normal body temperature can be nonpermissive, and thus one can envision that the virus is attenuated because of poor replication at normal body temperature or at a slightly raised temperature. However, for other attenuated viruses the nonpermissive temperature as measured in tissue culture is higher than that typically induced in vivo by an attenuated virus (Ray et al., 1995, Belshe and Hissom, 1982, Sakata and Nakayama, 2011). In these instances, the viruses are either more sensitive to slight increases in temperature in vivo, or there are other attenuating effects of the observed sequence changes that are expressed at the permissive temperature. Here we assessed the virulence of the Urabe ts variant, UrA9, and that of a non-ts Urabe virus, UrA5, and determined genetic and functional differences between these closely related viruses at permissive and non-permissive temperatures. The temperature sensitivity of the low passage stocks (p4) of the two viruses 3X plaque purified from the Urabe AM9 vaccine (Institute Pasteur Merieux) (Wright et al., 2000) was assessed by measuring growth of each virus in Vero cells at 37oC to 40oC using a standard plaque assay carried out at 37oC (Brown et al., 1996). The titer of UrA9 was significantly reduced at the higher temperature when cells were infected at low MOI while the titer of UrA5 was the same at both temperatures (Figure 1A). The temperature sensitive phenotype of A9 was retained when cells were infected at higher MOI for at least one complete replication cycle of input virus (24 hr pi) (Figure 1B). This is the time point used for RNA isolation in subsequent experiments. UrA5 replicated equally well at both temperatures at this MOI (data not shown). At both MOIs the ts phenotype was evident at early times pi, suggesting effects at early stages of the virus replication cycle. There were no differences in growth of the two viruses at permissive temperature, confirming earlier observations (Brown et al., 1996, Wright et al., 2000). The two viruses were compared in the rat neurovirulence test, where the degree of hydrocephalus induced by the virus is determined by measuring the cross sectional area of the brain occupied by the lateral ventricle excluding the cerebellum (Rubin et al., 2000, 2005). In this model, attenuated viruses display between 0 and 5% hydrocephalus, while wild type mumps viruses generally score between 15 and 25% (Sauder et al., 2006). UrA9 showed a significantly lower neurovirulence than UrA5, with a score of 1.9% compared to 7.8% (p <0.001, by t test, n = 31 for A9; n = 64 for A5), and is thus considered to be attenuated, while UrA5 has intermediate virulence. It has been reported that defective interfering particles or genomes (DIPs/DIGs) of L-Zagreb mumps generated by growth in Vero cells contribute to attenuation in the rat neurovirulence assay (Santak et al., 2015) although in later experiments, passage of plaque purified variants, some of which were attenuated, did not result in synthesis of DIPs/DIGs (Ivancic-Jelecki et al., 2016). To assess whether DIPs could be contributing to the attenuation of UrA9, we tested stocks of UrA5 and UrA9 used in vivo for their ability to interfere
with the growth of another Ur virus (Santak et al., 2015). We found no evidence of such activity in either Ur stock using this assay, nor did we observe evidence of substantial differences in DI content between the viruses according to the deep sequencing and analysis method used by Killip et al., 2010. We sequenced the genomes of each virus using the conventional Sanger method, which detects the consensus sequence in the stocks of the viruses, and subsequently genomes were analyzed by deep sequencing. The genome of Urabe mumps virus is 15384 nt in length, containing 7 genes: NP (nucleoprotein), P (encoding phosphoprotein, V, and putative I protein), M (matrix), F (fusion), SH (small hydrophobic), HN (hemagglutinin-neuraminidase) and L (polymerase). For conventional sequencing, 13 overlapping PCR products were generated from total RNA isolated from infected Vero cells as per Shah et al. (2009) and directly sequenced (Stemcore Laboratories, Ottawa, Canada). When compared with a consensus sequence derived from all banked Ur mumps sequences, this method identified 4 unique sites in UrA9 and 4 in UrA5. For deep sequencing, RNA was isolated directly from viral stocks at the same passage as used in other experiments. After digestion with micrococcal nuclease, extracted nucleic acids were fragmented by focused-ultrasonication (Covaris) and the NEBNext mRNA Library Prep Master Mix Set for Illumina (New England BioLabs) was used to produce libraries with paired end adaptors. These were amplified using 12 cycles of PCR with multiplex indexed primers and purified by magnetic beads (Agencourt AMPure PCR purification system, BeckmanCoulter) prior to sequencing (Illumina), and comparison to a reference sequence, Ur 87-1004, which was isolated from a patient who had received the AM9 vaccine (Brown et al., 1996) and which represents a consensus sequence for Ur viruses. Deep sequencing confirmed that the nt substitutions found by Sanger sequencing were dominant in the virus stocks, with <0.3% variation at each site. Thus each virus possessed 4 unique nt substitutions that were virtually invariant (Table 1). The 2 unique substitutions in the attenuated ts virus that altered aa sequences were at nt8015/HN468 and nt10601/L722; the 2 noncoding changes were at nt9307/L290, and at nt15289 within the 5’ UTR of the genome. The substitution in HN was previously reported (Wright et al., 2000) and had been noted in bulk Ur vaccine lots (Sauder et al., 2006), as well as in other strains of mumps virus, such as attenuated strains Enders and Rubini and wild type SBL-1 (Yates et al., 1996). The substitutions in UrA5 would create aa changes at P212 and at L169, and two noncoding changes were in the M and L ORFs (Table 1). There were 2 additional dominant coding changes in UrA9 compared to the consensus but which had been reported in another Ur isolate known to be virulent (Shah et al., 2009) (Table 1). Analysis of additional sequence heterogeneity showed that the general pattern of sequence variation across the genomes was similar for both virus stocks, mostly below 0.25% in frequency, with the exception of additional variation in the L ORF of UrA5 (Figure 2). Regions of slightly increased variation were observed in both viruses between nts 600-900 and 1350-1550 in the NP ORF, between nts 4100 – 4300 in the M ORF and 5138-5793 in the F ORF. As we do not know how prevalent any particular variant must be to affect the phenotype, we arbitrarily set 1% as a threshold frequency to further examine heterogeneity compared to the reference sequence. Excluding the sites shown in Table 1, UrA9 possessed 34 sites with frequencies of SNPs between 1% and 21%, while UrA5 had 30 sites with frequencies ranging from 1 to 37%. Eleven of these sites coincided between the viruses, with similar nt substitutions and frequencies. Each virus had a single variable residue in the 3’ genome end; nt18 in UrA9, and nt116 in UrA5, and up to 8 variable nts in the 5’ genome ends, 5 of which were the same in both viruses. All of the SNPs between nt 5138 and 5793 would result in coding changes in F (F198 to F409), and four of these sites were observed in both viruses. The frequencies of these SNPs ranged from 8 to 28% in UrA5, and from 2 to 21% in UrA9. UrA9 had 5 additional SNPs with frequencies of 1 to 2% and 1 SNP with a higher frequency of 7.5% in this region of F, which lies
between the two heptad repeat domains in F1 important for generating the coiled-coil structure of F (Liu et al., 2006). UrA9 also possessed a unique SNP of 13% heterogeneity at nt 1966, just upstream of the P/V translational start site. The virulent virus displayed more sequence heterogeneity in the L ORF, particularly between nt14101 and 15055 (L1889 to L2206). Thus the overall sequence heterogeneity and sites of heterogeneity are similar for both viruses, with the exception of the higher sequence variability in the F ORF of UrA9 and the L ORF of UrA5. As noted above, the higher variability in the 5’ genomes ends was not of such magnitude as to indicate the presence of trailer copyback DI particles (Killip et al., 2010). Because there was no specific pattern of sequence heterogeneity unique in the attenuated temperature sensitive virus, other than the additional SNPs in F, we reasoned that the dominant substitution from asp to tyr at L722 of UrA9 could contribute to the temperature sensitive phenotype. This is a non-conservative change within Domain III of viral RNA polymerases (RdRp) (Poch et al., 1990, Sleat et al., 1993). While this residue is not within sequences highly conserved in other strains of mumps or paramyxoviruses, L722 is only 56 aa upstream of the GDNQ motif important for catalytic activity of paramyxovirus and rhabdovirus RdRps (Sidhu et al., 1993, Sleat et al., 1993). Other changes in UrA9 L ORF that might be of significance were a substitution at nt 11810 resulting in an aa change at L1125 in Domain IV, present at a frequency of 12%, and nt 12253 (L1606) displaying 4% variation. This nt lies between Domains V and VI. To test whether the transcription function of UrA9 L is reduced at the non-permissive temperature, we inoculated cells at MOI 5.0 with each virus, incubated at 40oC or 37oC, and at various times post infection (pi), total RNA was isolated and relative amounts of two viral mRNAs were assessed by qPCR. The P gene is the second ORF to be transcribed, while the HN gene is the fifth ORF; transcription from this gene should represent the ability of the polymerase complex to remain associated with the template. At 16 and 24 hr pi, transcription from the UrA9 P gene was reduced at 40oC by a factor of 16 to 35 fold, while transcription from the HN gene was reduced by 3 to 19 fold (n=4) (Figure 3A & B). Transcription from UrA5 genes was similar at both temperatures. In all experiments, the titers of UrA9 at 40oC were at least 1 log reduced compared to titers at 37oC; the titers of UrA5 were the same at both temperatures. To confirm that this reduction in transcription affected viral protein synthesis, we examined the amount of HN or N protein by western blot using rabbit antibodies specific for individual proteins prepared by one of us (SR) after incubation of infected cells at 37oC and 40oC. A representative experiment showing a western blot probed with anti-HN is shown in Figure 4. The relative intensity of protein bands was measured in 3 experiments using GelQuant.NET software (biochemlabsolutions.com), and the average reduction in the intensity of UrA9 proteins at the higher temperature relative to 37oC was 5.8 + 2.1 fold after normalization to actin, whereas the reduction in UrA5 proteins at 40oC compared to 37oC was < 1 (p=0.037 by t test). As a measure of genome replication, production of the +ve sense replication intermediate (RI) was examined by RT/PCR using a virus specific primer for cDNA synthesis that hybridizes to the 3’ end of the RI but not to mRNA and generating a product encompassing the UTR. At 37oC, the desired product increased over time from 4 to 24 hr pi for both viruses (Figure 5A). At 24 hr pi the synthesis of UrA9 RI was reduced by 2.5 to 5 fold at 40oC compared to 37oC (n=3, Figure 5B), while the RI amplicon of UrA5 was the same at both temperatures. We tested UrA9 and UrA5 supernatants generated in these experiments for the presence of interfering activity. All tested supernatants inhibited growth of another Ur isolate by 2-4.5 fold, but there was no evidence of increased interfering activity in UrA9 samples over UrA5, or in samples generated at 40oC compared to 37oC (n=2, data not shown). Consistent with our findings, temperature sensitivity of HPIV3 has been associated with reduced transcription (Ray et al., 1995), and 2 of the 3 mutations in RdRP of cold adapted attenuated HPIV3
cp45 that contribute to ts and att phenotypes were within Domain III, while the 3rd mutation was in Domain IV (Skiadopoulos et al., 1998). In bovine PIV3 (bPIV3), a single mutation in Domain IV conferred both temperature sensitivity and attenuation (Haller et al., 2001), and mutations introduced into this domain of Sendai virus L resulted in temperature sensitivity of transcription and replication (Feller et al., 2000). However, none of these mutations corresponded to L722, L1125 or L1606. A coding change at L736 of another attenuated mumps strain, 88-1961, was shown to affect transcription in a non-ts manner, although in this instance transcription was enhanced rather than reduced compared to wt (Malik et al., 2007). Mutations at various residues within Domain III of Sendai virus RdRP have also been shown to reduce either transcription or replication without being linked to a ts phenotype (Smallwood et al., 2002). In our experiments, we observed no differences in polymerase activity between the two viruses at the permissive temperature. Other dominant substitutions we observed in the UrA9 polymerase were at L512, which is the final residue in the region between Domains I and Domains II, and L1085 is in Domain IV in a stretch of residues highly conserved in parainfluenza viruses and rhabdoviruses (Sidhu et al., 1993), but again not corresponding to any of the ts or attenuating mutations described in other paramyxoviruses. Both these changes are preferentially selected by passage in Vero cells (Sauder et al., 2006), and have been observed in non-ts, partially attenuated Ur variants (Shah et al., 2009), but this does not rule out the possibility that these residues contribute to the UrA9 phenotypes in the context of the changes/variation at the other sites unique to UrA9 L. The mutations the RdRp associated with attenuation of UrGw7, which was plaque purified from the vaccine at the same time as UrA9, were at L163 and L320, and unlike UrA9, this virus displayed reduced growth in Vero cells at 37oC compared to other Ur viruses (Wright et al., 2000). Another major substitution in the UrA9 virus that could influence temperature sensitivity and attenuation is the C → T change at nt 15289. This residue is located immediately 5’ of sequences that have been defined as conserved element II (CRII) of the antigenome promoter which is part of a bipartite promoter that drives synthesis of new genomes from the RI (Murphy et al., 1998, Keller et al., 2001). It has been suggested that CRII functions to initiate encapsidation of new genomes, or alternately, encapsidation results in formation of a turn in the nucleocapsid helix which brings CRII into proximity with the conserved promoter element (CRI) at the 3’ terminus of the RI, thus generating a binding site for the viral RNA polymerase (Murphy et al., 1998). At the non-permissive temperature, this substitution could affect encapsidation and helix formation or the functioning of the promoter for the production of genomes from the RI, but is unlikely to affect transcription or production of the RI. A G → T substitution at nt 18 in the 3’ genome end of UrA9, which is present at a frequency of 2.3%, is within promoter element I in the bipartite promoter that drives transcription and RI synthesis (Hoffman et al., 2000), and thus could be affecting these functions. The final major substitution we found in our virus, the glu to lys shift at HN468, is predicted to be on the globular head of HN (Crennell et al., 2000). While this change could affect binding or promotion of fusion at the non-permissive temperature, in our experiments binding and entry were conducted at 37oC before shifting cultures to the higher temperature. As well, we did not observe greater growth reduction at 40oC when cells were infected with low versus high MOI, when virus production would be less affected by cell to cell spread. A mutation at a nearby residue, HN aa466 has been associated with attenuation of strain 88-1961, and was shown to reduce sialic acid binding, neuraminidase activity and replication in Vero cells at 37oC (Malik et al., 2007). UrA9 replicated as well in Vero cells at 37oC as virulent Ur viruses, but displayed more rapid fusion compared to more virulent viruses (Brown et al., 1996, Wright et al., 2000). This increase in fusion could be a result of the high frequency of variants with sequence substitutions in F in UrA9, as well as the mutation at HN468.
In summary, we have determined that neuroattenuated ts UrA9 possesses 4 unique dominant sequence substitutions compared to a Ur consensus sequence; HN468, L722 in Domain III of the viral RdRp, a noncoding substitution at nt9307 (L290), and a substitution at nt 15289 in the UTR of the genome. The control virulent virus UrA5 also possesses 4 unique nt substitutions, resulting in aa changes at P212 and L169 and noncoding changes at nt3734 (M157) and nt14212 (L1925) compared to other banked Ur sequences. The extent of sequence heterogeneity was similar between the two viruses, but the UrA9 stock did possess elevated variation in F that was not observed in UrA5, and the UrA5 population was more variable at the 5’ end of the L ORF. The ts phenotype of UrA9 was associated with reduced transcription, protein accumulation and synthesis of the genome RI at the non-permissive temperature. As sequence heterogeneity in the UrA9 L gene was limited, we believe that the mutation at L722 and possibly the minor changes at L1125 and L1606 are contributing to this phenotype, although virus variants with sequence substitutions at both 3’ and 5’ genome ends of UrA9 could also play a role. We still do not understand how the ts phenotype relates to attenuation which, in the rat neurovirulence test, is associated with reduced replication of the virus in vivo (Rubin et al., 2000). We observed no reduction in UrA9 growth, transcription, protein expression or RI synthesis at the permissive temperature relative to the more virulent virus in Vero cells, nor did we detect the presence of DI particles that could explain attenuation. Similarly, the major changes in the sequence of UrA5 and the presence of variants with changes in UrA5 RdRp did not result in increased transcription or replication of this virus relative to the attenuated isolate, at least in vitro. While our results have not identified the mechanisms of attenuation or virulence of Ur mumps viruses, they do confirm that attenuation of the Ur mumps strains is associated with a small number of major sequence substitutions in one or both of the major surface glycoproteins, in the RdRp and at the 5’ antigenome promoter.
Acknowledgments We acknowledge funding from the National Sciences and Engineering Research Council of Canada (NSERC). We thank Erica Langley and Daniel Ngo for technical support with sequencing and western blots, Dr. M. Laassri for deep sequencing, and Dr. C.J Sauder for his review of the manuscript.
References 1. Belshe R.B. and Hissom F.K. 1982. Cold adaptation of parainfluenza virus type 3: induction of three phenotypic markers. J. Med. Virol. 10:235-242. 2. Brown, E.G., Dimock, K., Wright, K.E. 1996. The Urabe AM9 mumps vaccine is a mixture of viruses differing at amino acid 335 of the hemagglutinin-neuraminidase gene with one form associated with disease. J. Inf. Dis. 174:619-622. 3. Crennell, S., Takimoto T., Portner A., Taylor G. 2000 Crystal structure of the multifunctional paramyxovirus hemagglutinin-neuraminidase. Nat. Struct. Biol. 7: 1068-1074. 4. Crookshanks, F.K., Belshe, R.B. 1984. Evaluation of cold adapted and temperature sensitive mutants of parainfluenza virus type 3 in weanling hamsters. J. Med. Virol. 13, 243-249. 5. Feller, J.A., Smallwood, S., Horikami, S.M., Moyer, S.A. 2000. Mutations in conserved domains IV and VI of the large (L) subunit of the Sendai Virus RNA polymerase give a spectrum of defective RNA synthesis phenotypes. Virology 269:426-439.
6. Hall, S.K., Stokes, A., Tierney, E.L., London, W.T., Belshe, R.B., Newman, F.C., Murphy, B.R. 1992. Cold-passage human parainfluenza type 3 viruses contain ts and non-ts mutations leading to attenuation in rhesus monkeys. Virus Res. 22: 173-184. 7. Haller, A.A., MacPhail M., Mitiku M., Tang R.S. 2001. A single amino acid substitution in the viral polymerase creates a temperature-sensitive and attenuated recombinant bovine parainfluenza virus type 3. Virology 288: 342-350. 8. Hoffman, M.A., Banerjee, A.K. 2000. Precise mapping of the replication and transcription promoters of human parainfluenza virus type 3. Virology 269:201–211. 9. Ivancic-Jelecki, J., Forcic, D., Jagusic, M., Kostuic-Gulija, T., Mazuran R., Balija M.L., Isakov, O., Shomron, N. 2016. Influence of population diversity on neurovirulence potential of plaque purified LZagreb variants. Vaccine 34: 2383-2389. 10. Karron, R.A., Wright, P.F., Crowe, J.E. Jr. Clements-Mann, M.L., Thompson, J., Makhene M., Casey R., Murphy, B.R. 1997. Evaluation of two live, cold-passaged, temperature-sensitive respiratory syncytial virus vaccines in chimpanzees and in human adults, infants and children. J. Infect. Dis., 176: 1428-1436. 11. Keller M.A, Murphy S.K., Parks G.D. 2001. RNA replication from the Simian Virus 5 antigenomic promoter requires three sequence-dependent elements separated by sequence-independent spacer regions. J. Virol. 75:3993-3998. 12. Killip, M.J., Young, D.F., Gatherer D., Ross C.S., Short, J.A.L., Davison A.J., Goodbourn S., Randall, R.E. 2010. Deep sequencing analysis of defective genomes of parainfluenza virus 5 and their role in interferon induction. 13. Le Nouën, C., Brock, L.G., Luongo C., McCarty T., Yang L., Mehedi M., Winner, E., Mueller S., Collins P.L., Buchholz U.J., DiNapoli J.M. 2014. Attenuation of human respiratory syncytial virus by genome-scale codon-pair deoptimization. PNAS 111, 13169-13174. 14. Liu Y., Xu Y., Lou Z., Zhu J., Hu X., Gao G.F., Qui B., Rao Z., Tien P., 2006. Structural characterization of mumps virus fusion protein core. Biochem Biophys Res Comm 348: 916-922 15. Malik T., Wolbert C., Mauldin J., Sauder C., Carbone K.M., Rubin S.A. 2007. Functional consequences of attenuating mutations in the haemagglutinin-neuraminidase, fusion and polymerase proteins of a wild-type mumps virus strain. J. Gen. Virol. 88:2533-2541. 16. Murphy S.K., Ito Y., Parks G.D. 1998. A functional antigenomic promoter for the paramyxovirus Simian Virus 5 requires proper spacing between an essential internal segment and the 3’ terminus. J.Virol. 72, 10-19. 17. Poch, O., Blumberg, B.M., Bougueleret, L., Tordo, N. 1990. Sequence comparison of five polymerases (L proteins) of unsegmented negative-strand RNA viruses: theoretical assignment of functional domains.J. Gen. Virol. 71:1153-1162. 18. Ray, R., Meyer K., Newman F.K., Belshe R.B. 1995. Characterization of a live, attenuated human parainfluenza type 3 virus candidate vaccine. J. Virol. 69: 1959-1963. 19. Rubin, S.A., Pletnikov, M., Taffs, R., Snoy, P.J., Kobasa, D., Brown, E.G., Wright, K.E., Carbone, K.M. 2000. Evaluation of a neonatal rat model for prediction of mumps virus neurovirulence in humans. J. Virol. 74, 5382-5384. 20. Rubin, S.A., Afzal, M.S., Powell, C.L., Bentley, M.L. Auda, G.R., Taffs, R.E., Carbone, K.M. 2005. The rat-based neurovirulence safety test for the assessment of mumps virus neurovirulence in humans: an international collaborative study. J. Inf. Dis. 191, 1123-1128. 21. Saika, S., Kidokoro, M., Kubonoya, H., Ito, K., Ohkawa, T., Aoki, A., Nagata, N., Suzuki, K. 2006. Development and biological properties of a new live attenuated mumps vaccine. Comparative Immunology, Microbiology& Infectious Diseases 29 (2006) 89–99.
22. Sakata, M., and Nakayama, T. 2011. Protease and helicase domains are related to the temperature sensitivity of wild-type rubella viruses. Vaccine 29, 1107-1113. 23. Šantak, M., Markušić, M., Balija M.L., Kopač, S.K., Jug, R., Örvell, C., Tomac, T., Forčić, D. 2015. Accumulation of defective interfering viral particles in oly a few passages in Vero cells attenuates mumps virus neurovirulence. Microbes and Infection 17:228-236. 24. Sauder C.J., Vandenburgh K.M., Islow R.C., Malik T., Carbone K. M., Rubin, S.A. 2006. Changes in mumps virus neurovirulence phenotype associated with quasispecies heterogeneity. Virology 350, 4857. 25. Shah, D., Vidal, S., Link, M.A., Rubin, S.A., Wright, K.E. 2009. Identification of genetic mutations associated with attenuation and changes in tropisms of Urabe mumps virus. J. Med. Virol. 81, 130-138. 26. Sidhu, M.S., Menonna, J.P., Cook, S.C., Dowling, P.C., Udem, S.A. 1993. Canine distemper virus L gene: sequence and comparison with related viruses. Virology 193: 50-65. 27. Skiadopoulos M.H., Durbin, A.P., Tatem, J.M., Wu S-L., Paschalis M., Tao T., Collins, P.L., Murphy, B.R. 1998. Three amino acid substitutions in the L protein of the human parainfluenza virus type 3 cp45 live attenuated vaccine candidate contribute to its temperature-sensitive and attenuation phenotypes. J. Virol. 72: 1762-1768. 28. Skiadopoulos, M.H., Surman, S., Tatem, J.M., Paschalis, M., Wu, S-L., Udem, S.A., Durbin, A.P., Collins, P.L., Murphy, B.R. 1999. Identification of mutations contributing to the temperature-sensitive, cold-adapted, and attenuation phenotypes of live –attenuated cold-passage 45 (cp45) human parainfluenza virus 3 candidate vaccine. J. Virol. 73:1374-1381. 29. Sleat D.E., and Banerjee A.K. 1993. Transcriptional activity and mutational anlysis of recombinant vesicular stomatitis virus RNA polymerase. J.Virol. 67: 1334-1339. 30. Smallwood S., Hövel T., Neubert W.J., Moyer S.A. 2002. Different substitutions at conserved amino acids in Domains II and III in the Sendai L RNA polymerase protein inactivate viral RNA synthesis. Virology 304: 135-145. 31. Stokes, A., Tierney, E.L. Sarris, C.M., Murphy B.R., Hall, S.L. 1993. The complete nucleotide sequence of two cold-adapted, temperature-sensitive attenuated mutant vaccine viruses (cp12 and cp45) derived from the JS strain of human parainfluenza virus type 3 (PIV3). Virus Res. 30, 43-52. 32. Wright, K.E., Dimock, K., Brown, E.G. 2000. Biological characteristics of genetic variants of Urabe AM9 mumps vaccine virus. Virus Res. 67, 49-57. 33. Yates P.J., Afzal M. A., Minor P.D. 1996. Antigenic and genetic variation of the HN protein of mumps virus strains. J. Gen. Virol. 77:2491-2497.
Figure 1. Replication kinetics of UrA9 and UrA5 in Vero cells at permissive and non-permissive temperatures. Monolayers of Vero cells were inoculated with either UrA9 or UrA5 at MOI = 0.1 (A) or at MOI = 5 (B, UrA9, UrA5 not shown). After 1 hr at 37oC, the inoculum was removed, plates were incubated at 37°C or 40°C and supernatants were collected at indicated times for assessment of infectious virus by standard plaque assay. Values represent means + SD of n=6 for panel A, except for the 96 hr point where n=3; n=5 for panel B.
Figure 2. Frequency of single nucleotide polymorphisms (SNPs) of UrA9 and UrA5. RNA was isolated directly from stocks of each virus after micrococcal nuclease digestion, fragmented by focussedultrasonication and a cDNA library was generated using NEBNext mRNA Library Prep Master Mix Set for Illumina (New England BioLabs), and 12 cycles of PCR were carried out prior to purification and deep sequencing using MiSeq (Illumina). Raw sequencing reads were aligned against a reference Ur genome AF314560.1 using open access custom software (https://hive.biochemistry.gwu.edu/dna.cgi?cmd=main). Panel A shows frequencies of SNPs for each virus across the genome compared to the reference sequence; Panel B frequencies between 0 and 5% for the same data sets.
Figure 3. Transcription of A9 and A5 mRNAs at permissive and non-permissive temperatures. Equal numbers of Vero cells were inoculated at 37oC for 1 hr with MOI=5 of either UrA9 or UrA5, then incubated at 37°C or 40°C. At indicated times pi, total RNA was isolated, cDNA was generated with Oligo dT primers using equal volumes of RNA, and viral HN (Panel A) or P (Panel B) specific amplicons representing mRNA were measured by qPCR (BioRad iTaq SYBR Green Supermix with Rox). The resulting amplicons were normalized to background amplicons from mock infected cells and to actin mRNA prior to calculating fold reduction of viral mRNA at 40oC compared to 37oC (2-ΔΔCT). Results show a representative experiment of 3.
Figure 4. Accumulation of viral proteins at permissive and non-permissive temperatures. Equal numbers of Vero cells were inoculated at 37oC with MOI=5 of either virus, then incubated at 37°C or 40°C for indicated times. Lysates prepared from cell equivalents were probed with rabbit antibodies to mumps HN mixed with antibodies to actin as a loading control. The results show a representative experiment of 3.
Figure 5. UrA9 and UrA5 genome replication at permissive and non-permissive temperatures. Equal numbers of Vero cells were inoculated at 37oC with MOI=5 of either UrA9 or Ur A5, then incubated at 37°C or 40°C. Total RNA was isolated at indicated times, cDNA produced using a primer
that hybridizes only to the viral replication intermediate (RI) (5’CCAAGGGGAGAAAGTAAA3’), and the change in RI levels at 40oC compared to 37oC was determined for each virus by conventional PCR after normalization to mock samples. Panel A shows amplified RI products over time and Panel B shows fold reduction in RI from one representative experiment of 3 as measured by densitometry.
Table 1. Nucleotide and amino acid differences between A9, A5, Gw7, 1004-10/2, and 87-1004 Nucleotide/AA A9
A5
87-1004 (consensus)*
Gw7*
1004-10/2*
2611/P212
C/pro
A/gln
C/pro
C/pro
C/pro
2611/V
C/pro
A/pro
C/pro
C/pro
C/pro
3734/M157
G/lys
A/lys
G/lys
G/lys
G/lys
8015/ HN468
A/lys
G/glu
G/glu
G/glu
G/glu
8942/L169
A/ile
G/val
A/ile
A/ile
A/ile
9307/L290
C/gly
T/gly
T/gly
T/gly
T/gly
9972/ L512
T/phe
C/ser
C/ser
C/ser
T/phe
10601/ L722
T/tyr
G/asp
G/asp
G/asp
G/asp
11692/L1085
T/phe
G/leu
G/leu
G/leu
T/phe
14212/L1925
G/met
A/met
G/met
G/met
G/met
15289 (5’ genome trailer)
T
C
C
C
C
15328
G
G
A/G mix
A
G
Vero cells were inoculated with A9 or A5 at MOI=0.1, and incubated at 37oC. Total RNA was isolated at 48-72 h post infection, reverse transcribed and amplified with mumps specific primers (Shah et al., 2009). *GenBank accession numbers AF314560.1, FJ375177.1, FJ375178. Unique sequences in UrA5 and UrA9 are bolded, differences from consensus shared with other Ur isolates are in italics.
S0168-1702(16)30551-2 http://dx.doi.org/doi:10.1016/j.virusres.2016.10.003 VIRUS 96971
To appear in:
Virus Research
Received date: Revised date: Accepted date:
30-8-2016 4-10-2016 5-10-2016
Please cite this article as: Schinkel, Stephanie C.Burke, Rubin, Steven, Wright, Kathryn E., Mechanisms of temperature sensitivity of attenuated Urabe mumps virus.Virus Research http://dx.doi.org/10.1016/j.virusres.2016.10.003 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Mechanisms of temperature sensitivity of attenuated Urabe mumps virus. Stephanie C. Burke Schinkela , Steven Rubinb and Kathryn E. Wrighta* a
Department of Biochemistry, Microbiology and Immunology, University of Ottawa 451 Smyth Road, Ottawa, Canada K1H 8M5 b Food and Drug Administration (FDA), Center for Biologics Evaluation and Research (CBER), OVRR, DVP, 10903 New Hampshire Avenue, Bldg 52/72, Silver Spring, MD, 20993 [email protected] [email protected] [email protected] *corresponding author
Highlights
Attenuation of viruses is often linked to temperature sensitivity A temperature sensitive isolate of Urabe mumps virus was shown to be attenuated Genetic substitutions unique to this virus were identified including changes in the virus polymerase Functioning of the viral RNA polymerase was reduced at the non-permissive temperature
Abstract Temperature sensitivity is a phenotype often associated with attenuation of viruses. Previously, we purified several mumps variants from an incompletely attenuated Urabe strain live attenuated vaccine. Here we characterize one isolate that is sensitive to growth at high temperature. This virus was attenuated in a small animal model of mumps virulence, and we identified unique coding substitutions in the hemagglutinin-neuraminidase (HN), the viral polymerase (L) gene, and a non-coding substitution close to the anti-genome promoter sequences. At the non-permissive temperature, transcription of viral mRNAs and production of the replication intermediate were reduced compared to events at the permissive temperature and to a non-ts virulent Urabe virus. As well, synthesis of viral proteins was also reduced at the higher temperature. While the actual sequence substitutions in the ts virus were unique, the pattern of substitutions in HN, L and genome end sequences is similar to another attenuated Urabe virus previously described by us.
Keywords Mumps virus, attenuation, temperature sensitivity, transcription, replication
The genetic basis of virus attenuation (att) is complex in that each attenuated virus may have unique mutations, but attenuating mutations often have similar functional implications. We have shown that the AM9 Urabe (Ur) mumps vaccine, which was incompletely attenuated, contained several genetic variants of the Ur virus (Brown et al., 1996). Plaque purified isolates from the vaccine displayed differences in growth and fusion properties in tissue culture (Wright et al., 2000), displayed differences in attenuation in the rat neurovirulence model (Shah et al., 2009), and at least two viruses were temperature sensitive (ts). We had reasoned that the ts variants were likely amongst the most attenuated variants within the vaccine (Wright et al., 2000). Ts mutants of viruses commonly arise during attenuation processes such as serial passage of viruses in non-host cells (Saika et al., 2006, Sakata and Nakayama, 2011), codon-pair deoptimization (Le Nouën et al., 2014) and cold adaptation (Karron et al., 1997, Hall et al., 1992), and the ts phenotype is linked to attenuation in vivo (Crookshanks and Belshe, 1984). It is not clear how mutations attributing the ts phenotype also affect attenuation. For some viruses, particularly those that are also cold adapted (ca), normal body temperature can be nonpermissive, and thus one can envision that the virus is attenuated because of poor replication at normal body temperature or at a slightly raised temperature. However, for other attenuated viruses the nonpermissive temperature as measured in tissue culture is higher than that typically induced in vivo by an attenuated virus (Ray et al., 1995, Belshe and Hissom, 1982, Sakata and Nakayama, 2011). In these instances, the viruses are either more sensitive to slight increases in temperature in vivo, or there are other attenuating effects of the observed sequence changes that are expressed at the permissive temperature. Here we assessed the virulence of the Urabe ts variant, UrA9, and that of a non-ts Urabe virus, UrA5, and determined genetic and functional differences between these closely related viruses at permissive and non-permissive temperatures. The temperature sensitivity of the low passage stocks (p4) of the two viruses 3X plaque purified from the Urabe AM9 vaccine (Institute Pasteur Merieux) (Wright et al., 2000) was assessed by measuring growth of each virus in Vero cells at 37oC to 40oC using a standard plaque assay carried out at 37oC (Brown et al., 1996). The titer of UrA9 was significantly reduced at the higher temperature when cells were infected at low MOI while the titer of UrA5 was the same at both temperatures (Figure 1A). The temperature sensitive phenotype of A9 was retained when cells were infected at higher MOI for at least one complete replication cycle of input virus (24 hr pi) (Figure 1B). This is the time point used for RNA isolation in subsequent experiments. UrA5 replicated equally well at both temperatures at this MOI (data not shown). At both MOIs the ts phenotype was evident at early times pi, suggesting effects at early stages of the virus replication cycle. There were no differences in growth of the two viruses at permissive temperature, confirming earlier observations (Brown et al., 1996, Wright et al., 2000). The two viruses were compared in the rat neurovirulence test, where the degree of hydrocephalus induced by the virus is determined by measuring the cross sectional area of the brain occupied by the lateral ventricle excluding the cerebellum (Rubin et al., 2000, 2005). In this model, attenuated viruses display between 0 and 5% hydrocephalus, while wild type mumps viruses generally score between 15 and 25% (Sauder et al., 2006). UrA9 showed a significantly lower neurovirulence than UrA5, with a score of 1.9% compared to 7.8% (p <0.001, by t test, n = 31 for A9; n = 64 for A5), and is thus considered to be attenuated, while UrA5 has intermediate virulence. It has been reported that defective interfering particles or genomes (DIPs/DIGs) of L-Zagreb mumps generated by growth in Vero cells contribute to attenuation in the rat neurovirulence assay (Santak et al., 2015) although in later experiments, passage of plaque purified variants, some of which were attenuated, did not result in synthesis of DIPs/DIGs (Ivancic-Jelecki et al., 2016). To assess whether DIPs could be contributing to the attenuation of UrA9, we tested stocks of UrA5 and UrA9 used in vivo for their ability to interfere
with the growth of another Ur virus (Santak et al., 2015). We found no evidence of such activity in either Ur stock using this assay, nor did we observe evidence of substantial differences in DI content between the viruses according to the deep sequencing and analysis method used by Killip et al., 2010. We sequenced the genomes of each virus using the conventional Sanger method, which detects the consensus sequence in the stocks of the viruses, and subsequently genomes were analyzed by deep sequencing. The genome of Urabe mumps virus is 15384 nt in length, containing 7 genes: NP (nucleoprotein), P (encoding phosphoprotein, V, and putative I protein), M (matrix), F (fusion), SH (small hydrophobic), HN (hemagglutinin-neuraminidase) and L (polymerase). For conventional sequencing, 13 overlapping PCR products were generated from total RNA isolated from infected Vero cells as per Shah et al. (2009) and directly sequenced (Stemcore Laboratories, Ottawa, Canada). When compared with a consensus sequence derived from all banked Ur mumps sequences, this method identified 4 unique sites in UrA9 and 4 in UrA5. For deep sequencing, RNA was isolated directly from viral stocks at the same passage as used in other experiments. After digestion with micrococcal nuclease, extracted nucleic acids were fragmented by focused-ultrasonication (Covaris) and the NEBNext mRNA Library Prep Master Mix Set for Illumina (New England BioLabs) was used to produce libraries with paired end adaptors. These were amplified using 12 cycles of PCR with multiplex indexed primers and purified by magnetic beads (Agencourt AMPure PCR purification system, BeckmanCoulter) prior to sequencing (Illumina), and comparison to a reference sequence, Ur 87-1004, which was isolated from a patient who had received the AM9 vaccine (Brown et al., 1996) and which represents a consensus sequence for Ur viruses. Deep sequencing confirmed that the nt substitutions found by Sanger sequencing were dominant in the virus stocks, with <0.3% variation at each site. Thus each virus possessed 4 unique nt substitutions that were virtually invariant (Table 1). The 2 unique substitutions in the attenuated ts virus that altered aa sequences were at nt8015/HN468 and nt10601/L722; the 2 noncoding changes were at nt9307/L290, and at nt15289 within the 5’ UTR of the genome. The substitution in HN was previously reported (Wright et al., 2000) and had been noted in bulk Ur vaccine lots (Sauder et al., 2006), as well as in other strains of mumps virus, such as attenuated strains Enders and Rubini and wild type SBL-1 (Yates et al., 1996). The substitutions in UrA5 would create aa changes at P212 and at L169, and two noncoding changes were in the M and L ORFs (Table 1). There were 2 additional dominant coding changes in UrA9 compared to the consensus but which had been reported in another Ur isolate known to be virulent (Shah et al., 2009) (Table 1). Analysis of additional sequence heterogeneity showed that the general pattern of sequence variation across the genomes was similar for both virus stocks, mostly below 0.25% in frequency, with the exception of additional variation in the L ORF of UrA5 (Figure 2). Regions of slightly increased variation were observed in both viruses between nts 600-900 and 1350-1550 in the NP ORF, between nts 4100 – 4300 in the M ORF and 5138-5793 in the F ORF. As we do not know how prevalent any particular variant must be to affect the phenotype, we arbitrarily set 1% as a threshold frequency to further examine heterogeneity compared to the reference sequence. Excluding the sites shown in Table 1, UrA9 possessed 34 sites with frequencies of SNPs between 1% and 21%, while UrA5 had 30 sites with frequencies ranging from 1 to 37%. Eleven of these sites coincided between the viruses, with similar nt substitutions and frequencies. Each virus had a single variable residue in the 3’ genome end; nt18 in UrA9, and nt116 in UrA5, and up to 8 variable nts in the 5’ genome ends, 5 of which were the same in both viruses. All of the SNPs between nt 5138 and 5793 would result in coding changes in F (F198 to F409), and four of these sites were observed in both viruses. The frequencies of these SNPs ranged from 8 to 28% in UrA5, and from 2 to 21% in UrA9. UrA9 had 5 additional SNPs with frequencies of 1 to 2% and 1 SNP with a higher frequency of 7.5% in this region of F, which lies
between the two heptad repeat domains in F1 important for generating the coiled-coil structure of F (Liu et al., 2006). UrA9 also possessed a unique SNP of 13% heterogeneity at nt 1966, just upstream of the P/V translational start site. The virulent virus displayed more sequence heterogeneity in the L ORF, particularly between nt14101 and 15055 (L1889 to L2206). Thus the overall sequence heterogeneity and sites of heterogeneity are similar for both viruses, with the exception of the higher sequence variability in the F ORF of UrA9 and the L ORF of UrA5. As noted above, the higher variability in the 5’ genomes ends was not of such magnitude as to indicate the presence of trailer copyback DI particles (Killip et al., 2010). Because there was no specific pattern of sequence heterogeneity unique in the attenuated temperature sensitive virus, other than the additional SNPs in F, we reasoned that the dominant substitution from asp to tyr at L722 of UrA9 could contribute to the temperature sensitive phenotype. This is a non-conservative change within Domain III of viral RNA polymerases (RdRp) (Poch et al., 1990, Sleat et al., 1993). While this residue is not within sequences highly conserved in other strains of mumps or paramyxoviruses, L722 is only 56 aa upstream of the GDNQ motif important for catalytic activity of paramyxovirus and rhabdovirus RdRps (Sidhu et al., 1993, Sleat et al., 1993). Other changes in UrA9 L ORF that might be of significance were a substitution at nt 11810 resulting in an aa change at L1125 in Domain IV, present at a frequency of 12%, and nt 12253 (L1606) displaying 4% variation. This nt lies between Domains V and VI. To test whether the transcription function of UrA9 L is reduced at the non-permissive temperature, we inoculated cells at MOI 5.0 with each virus, incubated at 40oC or 37oC, and at various times post infection (pi), total RNA was isolated and relative amounts of two viral mRNAs were assessed by qPCR. The P gene is the second ORF to be transcribed, while the HN gene is the fifth ORF; transcription from this gene should represent the ability of the polymerase complex to remain associated with the template. At 16 and 24 hr pi, transcription from the UrA9 P gene was reduced at 40oC by a factor of 16 to 35 fold, while transcription from the HN gene was reduced by 3 to 19 fold (n=4) (Figure 3A & B). Transcription from UrA5 genes was similar at both temperatures. In all experiments, the titers of UrA9 at 40oC were at least 1 log reduced compared to titers at 37oC; the titers of UrA5 were the same at both temperatures. To confirm that this reduction in transcription affected viral protein synthesis, we examined the amount of HN or N protein by western blot using rabbit antibodies specific for individual proteins prepared by one of us (SR) after incubation of infected cells at 37oC and 40oC. A representative experiment showing a western blot probed with anti-HN is shown in Figure 4. The relative intensity of protein bands was measured in 3 experiments using GelQuant.NET software (biochemlabsolutions.com), and the average reduction in the intensity of UrA9 proteins at the higher temperature relative to 37oC was 5.8 + 2.1 fold after normalization to actin, whereas the reduction in UrA5 proteins at 40oC compared to 37oC was < 1 (p=0.037 by t test). As a measure of genome replication, production of the +ve sense replication intermediate (RI) was examined by RT/PCR using a virus specific primer for cDNA synthesis that hybridizes to the 3’ end of the RI but not to mRNA and generating a product encompassing the UTR. At 37oC, the desired product increased over time from 4 to 24 hr pi for both viruses (Figure 5A). At 24 hr pi the synthesis of UrA9 RI was reduced by 2.5 to 5 fold at 40oC compared to 37oC (n=3, Figure 5B), while the RI amplicon of UrA5 was the same at both temperatures. We tested UrA9 and UrA5 supernatants generated in these experiments for the presence of interfering activity. All tested supernatants inhibited growth of another Ur isolate by 2-4.5 fold, but there was no evidence of increased interfering activity in UrA9 samples over UrA5, or in samples generated at 40oC compared to 37oC (n=2, data not shown). Consistent with our findings, temperature sensitivity of HPIV3 has been associated with reduced transcription (Ray et al., 1995), and 2 of the 3 mutations in RdRP of cold adapted attenuated HPIV3
cp45 that contribute to ts and att phenotypes were within Domain III, while the 3rd mutation was in Domain IV (Skiadopoulos et al., 1998). In bovine PIV3 (bPIV3), a single mutation in Domain IV conferred both temperature sensitivity and attenuation (Haller et al., 2001), and mutations introduced into this domain of Sendai virus L resulted in temperature sensitivity of transcription and replication (Feller et al., 2000). However, none of these mutations corresponded to L722, L1125 or L1606. A coding change at L736 of another attenuated mumps strain, 88-1961, was shown to affect transcription in a non-ts manner, although in this instance transcription was enhanced rather than reduced compared to wt (Malik et al., 2007). Mutations at various residues within Domain III of Sendai virus RdRP have also been shown to reduce either transcription or replication without being linked to a ts phenotype (Smallwood et al., 2002). In our experiments, we observed no differences in polymerase activity between the two viruses at the permissive temperature. Other dominant substitutions we observed in the UrA9 polymerase were at L512, which is the final residue in the region between Domains I and Domains II, and L1085 is in Domain IV in a stretch of residues highly conserved in parainfluenza viruses and rhabdoviruses (Sidhu et al., 1993), but again not corresponding to any of the ts or attenuating mutations described in other paramyxoviruses. Both these changes are preferentially selected by passage in Vero cells (Sauder et al., 2006), and have been observed in non-ts, partially attenuated Ur variants (Shah et al., 2009), but this does not rule out the possibility that these residues contribute to the UrA9 phenotypes in the context of the changes/variation at the other sites unique to UrA9 L. The mutations the RdRp associated with attenuation of UrGw7, which was plaque purified from the vaccine at the same time as UrA9, were at L163 and L320, and unlike UrA9, this virus displayed reduced growth in Vero cells at 37oC compared to other Ur viruses (Wright et al., 2000). Another major substitution in the UrA9 virus that could influence temperature sensitivity and attenuation is the C → T change at nt 15289. This residue is located immediately 5’ of sequences that have been defined as conserved element II (CRII) of the antigenome promoter which is part of a bipartite promoter that drives synthesis of new genomes from the RI (Murphy et al., 1998, Keller et al., 2001). It has been suggested that CRII functions to initiate encapsidation of new genomes, or alternately, encapsidation results in formation of a turn in the nucleocapsid helix which brings CRII into proximity with the conserved promoter element (CRI) at the 3’ terminus of the RI, thus generating a binding site for the viral RNA polymerase (Murphy et al., 1998). At the non-permissive temperature, this substitution could affect encapsidation and helix formation or the functioning of the promoter for the production of genomes from the RI, but is unlikely to affect transcription or production of the RI. A G → T substitution at nt 18 in the 3’ genome end of UrA9, which is present at a frequency of 2.3%, is within promoter element I in the bipartite promoter that drives transcription and RI synthesis (Hoffman et al., 2000), and thus could be affecting these functions. The final major substitution we found in our virus, the glu to lys shift at HN468, is predicted to be on the globular head of HN (Crennell et al., 2000). While this change could affect binding or promotion of fusion at the non-permissive temperature, in our experiments binding and entry were conducted at 37oC before shifting cultures to the higher temperature. As well, we did not observe greater growth reduction at 40oC when cells were infected with low versus high MOI, when virus production would be less affected by cell to cell spread. A mutation at a nearby residue, HN aa466 has been associated with attenuation of strain 88-1961, and was shown to reduce sialic acid binding, neuraminidase activity and replication in Vero cells at 37oC (Malik et al., 2007). UrA9 replicated as well in Vero cells at 37oC as virulent Ur viruses, but displayed more rapid fusion compared to more virulent viruses (Brown et al., 1996, Wright et al., 2000). This increase in fusion could be a result of the high frequency of variants with sequence substitutions in F in UrA9, as well as the mutation at HN468.
In summary, we have determined that neuroattenuated ts UrA9 possesses 4 unique dominant sequence substitutions compared to a Ur consensus sequence; HN468, L722 in Domain III of the viral RdRp, a noncoding substitution at nt9307 (L290), and a substitution at nt 15289 in the UTR of the genome. The control virulent virus UrA5 also possesses 4 unique nt substitutions, resulting in aa changes at P212 and L169 and noncoding changes at nt3734 (M157) and nt14212 (L1925) compared to other banked Ur sequences. The extent of sequence heterogeneity was similar between the two viruses, but the UrA9 stock did possess elevated variation in F that was not observed in UrA5, and the UrA5 population was more variable at the 5’ end of the L ORF. The ts phenotype of UrA9 was associated with reduced transcription, protein accumulation and synthesis of the genome RI at the non-permissive temperature. As sequence heterogeneity in the UrA9 L gene was limited, we believe that the mutation at L722 and possibly the minor changes at L1125 and L1606 are contributing to this phenotype, although virus variants with sequence substitutions at both 3’ and 5’ genome ends of UrA9 could also play a role. We still do not understand how the ts phenotype relates to attenuation which, in the rat neurovirulence test, is associated with reduced replication of the virus in vivo (Rubin et al., 2000). We observed no reduction in UrA9 growth, transcription, protein expression or RI synthesis at the permissive temperature relative to the more virulent virus in Vero cells, nor did we detect the presence of DI particles that could explain attenuation. Similarly, the major changes in the sequence of UrA5 and the presence of variants with changes in UrA5 RdRp did not result in increased transcription or replication of this virus relative to the attenuated isolate, at least in vitro. While our results have not identified the mechanisms of attenuation or virulence of Ur mumps viruses, they do confirm that attenuation of the Ur mumps strains is associated with a small number of major sequence substitutions in one or both of the major surface glycoproteins, in the RdRp and at the 5’ antigenome promoter.
Acknowledgments We acknowledge funding from the National Sciences and Engineering Research Council of Canada (NSERC). We thank Erica Langley and Daniel Ngo for technical support with sequencing and western blots, Dr. M. Laassri for deep sequencing, and Dr. C.J Sauder for his review of the manuscript.
References 1. Belshe R.B. and Hissom F.K. 1982. Cold adaptation of parainfluenza virus type 3: induction of three phenotypic markers. J. Med. Virol. 10:235-242. 2. Brown, E.G., Dimock, K., Wright, K.E. 1996. The Urabe AM9 mumps vaccine is a mixture of viruses differing at amino acid 335 of the hemagglutinin-neuraminidase gene with one form associated with disease. J. Inf. Dis. 174:619-622. 3. Crennell, S., Takimoto T., Portner A., Taylor G. 2000 Crystal structure of the multifunctional paramyxovirus hemagglutinin-neuraminidase. Nat. Struct. Biol. 7: 1068-1074. 4. Crookshanks, F.K., Belshe, R.B. 1984. Evaluation of cold adapted and temperature sensitive mutants of parainfluenza virus type 3 in weanling hamsters. J. Med. Virol. 13, 243-249. 5. Feller, J.A., Smallwood, S., Horikami, S.M., Moyer, S.A. 2000. Mutations in conserved domains IV and VI of the large (L) subunit of the Sendai Virus RNA polymerase give a spectrum of defective RNA synthesis phenotypes. Virology 269:426-439.
6. Hall, S.K., Stokes, A., Tierney, E.L., London, W.T., Belshe, R.B., Newman, F.C., Murphy, B.R. 1992. Cold-passage human parainfluenza type 3 viruses contain ts and non-ts mutations leading to attenuation in rhesus monkeys. Virus Res. 22: 173-184. 7. Haller, A.A., MacPhail M., Mitiku M., Tang R.S. 2001. A single amino acid substitution in the viral polymerase creates a temperature-sensitive and attenuated recombinant bovine parainfluenza virus type 3. Virology 288: 342-350. 8. Hoffman, M.A., Banerjee, A.K. 2000. Precise mapping of the replication and transcription promoters of human parainfluenza virus type 3. Virology 269:201–211. 9. Ivancic-Jelecki, J., Forcic, D., Jagusic, M., Kostuic-Gulija, T., Mazuran R., Balija M.L., Isakov, O., Shomron, N. 2016. Influence of population diversity on neurovirulence potential of plaque purified LZagreb variants. Vaccine 34: 2383-2389. 10. Karron, R.A., Wright, P.F., Crowe, J.E. Jr. Clements-Mann, M.L., Thompson, J., Makhene M., Casey R., Murphy, B.R. 1997. Evaluation of two live, cold-passaged, temperature-sensitive respiratory syncytial virus vaccines in chimpanzees and in human adults, infants and children. J. Infect. Dis., 176: 1428-1436. 11. Keller M.A, Murphy S.K., Parks G.D. 2001. RNA replication from the Simian Virus 5 antigenomic promoter requires three sequence-dependent elements separated by sequence-independent spacer regions. J. Virol. 75:3993-3998. 12. Killip, M.J., Young, D.F., Gatherer D., Ross C.S., Short, J.A.L., Davison A.J., Goodbourn S., Randall, R.E. 2010. Deep sequencing analysis of defective genomes of parainfluenza virus 5 and their role in interferon induction. 13. Le Nouën, C., Brock, L.G., Luongo C., McCarty T., Yang L., Mehedi M., Winner, E., Mueller S., Collins P.L., Buchholz U.J., DiNapoli J.M. 2014. Attenuation of human respiratory syncytial virus by genome-scale codon-pair deoptimization. PNAS 111, 13169-13174. 14. Liu Y., Xu Y., Lou Z., Zhu J., Hu X., Gao G.F., Qui B., Rao Z., Tien P., 2006. Structural characterization of mumps virus fusion protein core. Biochem Biophys Res Comm 348: 916-922 15. Malik T., Wolbert C., Mauldin J., Sauder C., Carbone K.M., Rubin S.A. 2007. Functional consequences of attenuating mutations in the haemagglutinin-neuraminidase, fusion and polymerase proteins of a wild-type mumps virus strain. J. Gen. Virol. 88:2533-2541. 16. Murphy S.K., Ito Y., Parks G.D. 1998. A functional antigenomic promoter for the paramyxovirus Simian Virus 5 requires proper spacing between an essential internal segment and the 3’ terminus. J.Virol. 72, 10-19. 17. Poch, O., Blumberg, B.M., Bougueleret, L., Tordo, N. 1990. Sequence comparison of five polymerases (L proteins) of unsegmented negative-strand RNA viruses: theoretical assignment of functional domains.J. Gen. Virol. 71:1153-1162. 18. Ray, R., Meyer K., Newman F.K., Belshe R.B. 1995. Characterization of a live, attenuated human parainfluenza type 3 virus candidate vaccine. J. Virol. 69: 1959-1963. 19. Rubin, S.A., Pletnikov, M., Taffs, R., Snoy, P.J., Kobasa, D., Brown, E.G., Wright, K.E., Carbone, K.M. 2000. Evaluation of a neonatal rat model for prediction of mumps virus neurovirulence in humans. J. Virol. 74, 5382-5384. 20. Rubin, S.A., Afzal, M.S., Powell, C.L., Bentley, M.L. Auda, G.R., Taffs, R.E., Carbone, K.M. 2005. The rat-based neurovirulence safety test for the assessment of mumps virus neurovirulence in humans: an international collaborative study. J. Inf. Dis. 191, 1123-1128. 21. Saika, S., Kidokoro, M., Kubonoya, H., Ito, K., Ohkawa, T., Aoki, A., Nagata, N., Suzuki, K. 2006. Development and biological properties of a new live attenuated mumps vaccine. Comparative Immunology, Microbiology& Infectious Diseases 29 (2006) 89–99.
22. Sakata, M., and Nakayama, T. 2011. Protease and helicase domains are related to the temperature sensitivity of wild-type rubella viruses. Vaccine 29, 1107-1113. 23. Šantak, M., Markušić, M., Balija M.L., Kopač, S.K., Jug, R., Örvell, C., Tomac, T., Forčić, D. 2015. Accumulation of defective interfering viral particles in oly a few passages in Vero cells attenuates mumps virus neurovirulence. Microbes and Infection 17:228-236. 24. Sauder C.J., Vandenburgh K.M., Islow R.C., Malik T., Carbone K. M., Rubin, S.A. 2006. Changes in mumps virus neurovirulence phenotype associated with quasispecies heterogeneity. Virology 350, 4857. 25. Shah, D., Vidal, S., Link, M.A., Rubin, S.A., Wright, K.E. 2009. Identification of genetic mutations associated with attenuation and changes in tropisms of Urabe mumps virus. J. Med. Virol. 81, 130-138. 26. Sidhu, M.S., Menonna, J.P., Cook, S.C., Dowling, P.C., Udem, S.A. 1993. Canine distemper virus L gene: sequence and comparison with related viruses. Virology 193: 50-65. 27. Skiadopoulos M.H., Durbin, A.P., Tatem, J.M., Wu S-L., Paschalis M., Tao T., Collins, P.L., Murphy, B.R. 1998. Three amino acid substitutions in the L protein of the human parainfluenza virus type 3 cp45 live attenuated vaccine candidate contribute to its temperature-sensitive and attenuation phenotypes. J. Virol. 72: 1762-1768. 28. Skiadopoulos, M.H., Surman, S., Tatem, J.M., Paschalis, M., Wu, S-L., Udem, S.A., Durbin, A.P., Collins, P.L., Murphy, B.R. 1999. Identification of mutations contributing to the temperature-sensitive, cold-adapted, and attenuation phenotypes of live –attenuated cold-passage 45 (cp45) human parainfluenza virus 3 candidate vaccine. J. Virol. 73:1374-1381. 29. Sleat D.E., and Banerjee A.K. 1993. Transcriptional activity and mutational anlysis of recombinant vesicular stomatitis virus RNA polymerase. J.Virol. 67: 1334-1339. 30. Smallwood S., Hövel T., Neubert W.J., Moyer S.A. 2002. Different substitutions at conserved amino acids in Domains II and III in the Sendai L RNA polymerase protein inactivate viral RNA synthesis. Virology 304: 135-145. 31. Stokes, A., Tierney, E.L. Sarris, C.M., Murphy B.R., Hall, S.L. 1993. The complete nucleotide sequence of two cold-adapted, temperature-sensitive attenuated mutant vaccine viruses (cp12 and cp45) derived from the JS strain of human parainfluenza virus type 3 (PIV3). Virus Res. 30, 43-52. 32. Wright, K.E., Dimock, K., Brown, E.G. 2000. Biological characteristics of genetic variants of Urabe AM9 mumps vaccine virus. Virus Res. 67, 49-57. 33. Yates P.J., Afzal M. A., Minor P.D. 1996. Antigenic and genetic variation of the HN protein of mumps virus strains. J. Gen. Virol. 77:2491-2497.
Figure 1. Replication kinetics of UrA9 and UrA5 in Vero cells at permissive and non-permissive temperatures. Monolayers of Vero cells were inoculated with either UrA9 or UrA5 at MOI = 0.1 (A) or at MOI = 5 (B, UrA9, UrA5 not shown). After 1 hr at 37oC, the inoculum was removed, plates were incubated at 37°C or 40°C and supernatants were collected at indicated times for assessment of infectious virus by standard plaque assay. Values represent means + SD of n=6 for panel A, except for the 96 hr point where n=3; n=5 for panel B.
Figure 2. Frequency of single nucleotide polymorphisms (SNPs) of UrA9 and UrA5. RNA was isolated directly from stocks of each virus after micrococcal nuclease digestion, fragmented by focussedultrasonication and a cDNA library was generated using NEBNext mRNA Library Prep Master Mix Set for Illumina (New England BioLabs), and 12 cycles of PCR were carried out prior to purification and deep sequencing using MiSeq (Illumina). Raw sequencing reads were aligned against a reference Ur genome AF314560.1 using open access custom software (https://hive.biochemistry.gwu.edu/dna.cgi?cmd=main). Panel A shows frequencies of SNPs for each virus across the genome compared to the reference sequence; Panel B frequencies between 0 and 5% for the same data sets.
Figure 3. Transcription of A9 and A5 mRNAs at permissive and non-permissive temperatures. Equal numbers of Vero cells were inoculated at 37oC for 1 hr with MOI=5 of either UrA9 or UrA5, then incubated at 37°C or 40°C. At indicated times pi, total RNA was isolated, cDNA was generated with Oligo dT primers using equal volumes of RNA, and viral HN (Panel A) or P (Panel B) specific amplicons representing mRNA were measured by qPCR (BioRad iTaq SYBR Green Supermix with Rox). The resulting amplicons were normalized to background amplicons from mock infected cells and to actin mRNA prior to calculating fold reduction of viral mRNA at 40oC compared to 37oC (2-ΔΔCT). Results show a representative experiment of 3.
Figure 4. Accumulation of viral proteins at permissive and non-permissive temperatures. Equal numbers of Vero cells were inoculated at 37oC with MOI=5 of either virus, then incubated at 37°C or 40°C for indicated times. Lysates prepared from cell equivalents were probed with rabbit antibodies to mumps HN mixed with antibodies to actin as a loading control. The results show a representative experiment of 3.
Figure 5. UrA9 and UrA5 genome replication at permissive and non-permissive temperatures. Equal numbers of Vero cells were inoculated at 37oC with MOI=5 of either UrA9 or Ur A5, then incubated at 37°C or 40°C. Total RNA was isolated at indicated times, cDNA produced using a primer
that hybridizes only to the viral replication intermediate (RI) (5’CCAAGGGGAGAAAGTAAA3’), and the change in RI levels at 40oC compared to 37oC was determined for each virus by conventional PCR after normalization to mock samples. Panel A shows amplified RI products over time and Panel B shows fold reduction in RI from one representative experiment of 3 as measured by densitometry.
Table 1. Nucleotide and amino acid differences between A9, A5, Gw7, 1004-10/2, and 87-1004 Nucleotide/AA A9
A5
87-1004 (consensus)*
Gw7*
1004-10/2*
2611/P212
C/pro
A/gln
C/pro
C/pro
C/pro
2611/V
C/pro
A/pro
C/pro
C/pro
C/pro
3734/M157
G/lys
A/lys
G/lys
G/lys
G/lys
8015/ HN468
A/lys
G/glu
G/glu
G/glu
G/glu
8942/L169
A/ile
G/val
A/ile
A/ile
A/ile
9307/L290
C/gly
T/gly
T/gly
T/gly
T/gly
9972/ L512
T/phe
C/ser
C/ser
C/ser
T/phe
10601/ L722
T/tyr
G/asp
G/asp
G/asp
G/asp
11692/L1085
T/phe
G/leu
G/leu
G/leu
T/phe
14212/L1925
G/met
A/met
G/met
G/met
G/met
15289 (5’ genome trailer)
T
C
C
C
C
15328
G
G
A/G mix
A
G
Vero cells were inoculated with A9 or A5 at MOI=0.1, and incubated at 37oC. Total RNA was isolated at 48-72 h post infection, reverse transcribed and amplified with mumps specific primers (Shah et al., 2009). *GenBank accession numbers AF314560.1, FJ375177.1, FJ375178. Unique sequences in UrA5 and UrA9 are bolded, differences from consensus shared with other Ur isolates are in italics.