Highlights
- •The hemagglutinin–neuraminidase (HN) and small hydrophobic (SH) proteins of American and Dutch outbreak lineages were aligned.
- •Immunogenicity predictions indicated diversity in epitopes between these and Jeryl-Lynn (JL).
- •The literature describing serum neutralization in the context of these data was reviewed.
- •Heterologous epitopes and waning immunity have likely contributed to recent outbreaks.
- •Outbreaks in vaccinated patients invite consideration of a polyvalent mumps vaccine.
Abstract
Mumps outbreaks among vaccinated patients have become increasingly common in recent years. While there are multiple conditions driving this re-emergence, convention has suggested that these outbreaks are associated with waning immunity rather than vaccine escape. Molecular evidence from both the ongoing American and Dutch outbreaks in conjunction with recent structural biology studies challenge this convention, and suggest that emergent lineages of mumps virus exhibit key differences in antigenic epitopes from the vaccine strain employed: Jeryl-Lynn 5. The American and Dutch 2016–2017 outbreak lineages were examined using computational biology through the lens of diversity in immunogenic epitopes. Findings are discussed and the laboratory evidence indicating neutralization of heterologous mumps strains by serum from vaccinated individuals is reviewed. Taken together, it is concluded that the number of heterologous epitopes occurring in mumps virus in conjunction with waning immunity is facilitating small outbreaks in vaccinated patients, and that consideration of a polyvalent mumps vaccine is warranted.
Keywords
Mumps virus (MuV) is a member of the Paramyxoviridae and a cause of fever and viral parotitis. Less frequently, it is associated with orchitis, encephalitis, aseptic meningitis, deafness, and pancreatitis (
Rubin et al., 2015
). Routine immunization against mumps is part of the recommended series for children in the USA and Europe, and is given in combination with measles and rubella immunization (MMR) (M-M-R II; Merck & Co., West Point, PA, USA) or as measles, rubella, and varicella zoster immunization (ProQuad; Merck & Co., West Point, PA, USA). MuV exhibits moderate genetic heterogeneity (Ivancic-Jelecki et al., 2008
), and 12 genotypes (A–N, excluding E and M) currently or historically circulate in different parts of the world (Jin et al., 2015
). Multiple vaccine strains have been used in different locations; however, the discussion here is restricted to the genotype A vaccine strains used in Western Europe and the USA: Jeryl-Lynn (JL) lineages 2 and 5.Increased instances of mumps outbreaks in vaccinated individuals have been seen in recent years, with those on college campuses disproportionately affected (
Albertson et al., 2016
, Cortese et al., 2008
, Patel et al., 2017
). Reasons for this increase are likely multifaceted, and include declining levels of vaccine-derived immunity (LeBaron et al., 2009
, Davidkin et al., 2008
, Gu et al., 2017
) and a significant reduction in the natural ‘boosters’ received by vaccinated individuals as MuV has become less prevalent in the increasingly vaccinated population. The possibility of gradual immune escape by MuV variants has been discussed previously (- Gu X.X.
- Plotkin S.A.
- Edwards K.M.
- Sette A.
- Mills K.H.G.
- Levy O.
- et al.
Waning Immunity and Microbial Vaccines — Workshop of the National Institute of Allergy and Infectious Diseases.
Clin Vaccine Immunol. 2017; https://doi.org/10.1128/CVI.00034-17
Nöjd et al., 2001
), but was confounded by laboratory findings demonstrating neutralization of wild-type MuV strains representing diverse genotypes by serum from JL-vaccinated children (Gouma et al., 2016a
, Rubin et al., 2012
). The decrease in antibody titers over time following vaccination with live, attenuated viruses is not unique to mumps; however, the recent tendency toward outbreaks in vaccinated individuals seems to be.This prompted the present authors to (1) carefully examine the literature describing protective MuV epitopes, neutralization of heterologous strains by JL hyperimmune sera, and the natural history of MuV, and (2) perform a computational analysis of the emergent outbreak strains currently circulating in the Netherlands and the USA, in order to fully explore the potential for MuV vaccine escape.
Of the nine MuV proteins, the most thoroughly explored are the hemagglutinin–neuraminidase (HN), the fusion protein (F), and the small hydrophobic protein (SH). The SH protein exhibits elevated levels of diversity relative to the rest of the MuV genome, and for this reason it is frequently used for genotyping and epidemiological surveillance (
Orvell et al., 1997
; Ströhle et al., 1996
; Yeo et al., 1993
). Sera from vaccinated individuals and convalescent patients indicate that the immunodominant antigens are HN and F (Cusi et al., 2001
, Kulkarni-Kale et al., 2007
, Šantak et al., 2015
).The robustness of in vitro serum neutralization assays against HN and F were measured by generating chimeric MuV constructs wherein JL and wild-type genotype H strain 88-1961 backbones were supplemented with the reciprocal copy of HN and F. JL bearing 88-1961 F was neutralized by hyperimmune sera at equivalent levels to native JL, indicating that vaccine-derived antibodies against F are fully cross-reactive (
Rubin et al., 2012
). Neutralization of JL bearing the 88-1961 HN protein by JL hyperimmune serum was significantly different from JL and the JL/88-1961 F chimera, indicating that biologically relevant distinctions existed between at least some immunogenic epitopes in HN (Rubin et al., 2012
). Similarly, cross-neutralization between strains was full in strains featuring no amino acid changes from the relevant vaccine strain, and partial in strains where changes in protein sequence were apparent (Gouma et al., 2016a
, Vaidya et al., 2016
). These findings are suggestive of retention of certain HN epitopes across MuV strains and variation in others, but the threshold between lower capacity for neutralization in vitro translating into protection versus failure in vivo is unclear. Importantly, Rubin et al. note that the neutralizing sera used in their study were collected 6 weeks post-vaccination, and that antibody titers are not likely to be representative of the infected, vaccinated patients involved in recent outbreaks (Rubin et al., 2012
). The investigators also note that T-cell involvement in protective immunity was not assessed, which Homen and Bremel described as likely to be important (Homan and Bremel, 2014
).The lower capacity for neutralization of HN across heterologous strains may very well mean that vaccine-induced immunity drops below the threshold required for in vivo protection in the face of declining antibody titers. This scenario reflects the current state of mumps outbreaks, but would be bolstered by a computational examination of immunogenic protein sequences from outbreak strains in vaccinated populations for validation. Such an analysis is directly possible with publicly available MuV surveillance data from the Netherlands (
Gouma et al., 2016b
), and indirectly with surveillance data from the USA (Benson et al., 2017
). The crystal structure of MuV HN was recently described by Kubota et al., who noted that some of the previously noted three-dimensional immunogenic epitopes involve residues that are mutable across strains (Kubota et al., 2016
).In the present study, multiple sequence alignments were assembled using the HN amino acid sequence from the outbreak strains in the Netherlands for the years 2013–2015 and from JL, the vaccine strain used in the country. The outbreak strains are largely clonal from the standpoint of HN; more than 60 of the >100 isolates available in GenBank collected in the Netherlands over that time period share 100% amino acid identity, and another 16 differ by only a single substitution of a similar residue (lysine to arginine). Twenty Dutch isolates with sequences available in GenBank were selected and their sequence diversity examined (
Benson et al., 2017
, Sievers and Higgins, 2014
; dataset available as Supplementary material supplemental dataset S1 and at doi: 10.13140/RG.2.2.28761.42088), antigenic epitopes predicted (via http://imed.med.ucm.es/Tools/antigenic.html), and secondary structure computed (Garnier et al., 1996
) to address the probability of immune escape in a predictive way. The findings were then extrapolated to SH sequences from current USA outbreak strains.Only five of the 19 Dutch strains examined contained amino acid substitutions relative to consensus (10 total substitutions), indicating that these strains are highly homologous. While all but one of the changes were in predicted antigenic epitopes, only one generated a predicted change in secondary structure (Table 1). This alanine to valine change is conserved in two isolates collected in the Spring of 2015, and generates a predicted change from an α-helix to a linear β-strand. This suggests the loss of an antigenic epitope rather than the introduction of a novel one. These results indicate that the level of amino acid diversity is extremely low within the collection of Dutch outbreak strains. That said, comparison of their sequence, antigenic, and structural patterns to those of JL demonstrated a marked diversity between the strain generating immunity and the strains causing disease. This divergence was not uniform across the HN protein. Of the 25 predicted antigenic epitopes, 13 were 100% identical in sequence, antigenicity, and secondary structure between outbreak strains and JL. These conserved epitopes likely explain the observation of some neutralization of heterologous strains by JL hyperimmune serum (
Gouma et al., 2016a
, Rubin et al., 2012
, Vaidya et al., 2016
).Table 1Hemagglutinin–neuraminidase (HN) diversity among outbreak strains.
| Divergent strain | Number of sites | Divergent site | Structural change | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Leiden 15 | 5 | A13S | N25D | A37V | T130S | S462T | No | No | H → E | No | No |
| Heerhugowaard 13 | 1 | K317R | No | ||||||||
| Hilversum 14 | 1 | K317R | No | ||||||||
| Purmerend 14 | 2 | T97S | K317R | No | No | ||||||
| Hengevelde 15 | 2 | N25D | A37V | No | H → E | ||||||
a Divergence indicates difference from the consensus HN sequence of 19 Dutch outbreak strains examined. No small hydrophobic protein (SH) sequences featured changes in amino acid sequence across all American outbreak strains; therefore, SH diversity is not included in this table.
b Divergent sites are described by the amino acid of the consensus sequence, followed by the residue number, followed by the amino acid of the noted strain.
c Secondary structures are abbreviated as follows: H = helix; E = beta strand; C = coiled coil.
The remaining 12 epitopes all feature at least one divergent residue from JL among the outbreak strains, and 11 of these result in a predicted structural change (Table 2). It is notable that the number of divergent sites often generated a disproportionately large number of predicted structural changes, indicating that the impact of MuV diversity may be underestimated simply by tabulating the number of changes. This structural diversity suggests that antibodies raised against the JL epitopes may no longer interact with the analogous outbreak strain epitopes. Two of three divergent epitopes found in a recent genotype G isolate that facilitated loss of neutralization by sera from JL-immunized guinea pigs were identified in the computational analysis, validating the approach (
Šantak et al., 2013
). The tendency of vaccine-derived antibody titers against MuV to decline over time (LeBaron et al., 2009
) makes it is highly plausible that the recent Netherlands outbreaks in vaccinated individuals resulted from antibodies against the conserved epitopes falling below the protective threshold given their small number.Table 2Divergence between vaccine strain JL5 and outbreak strains.
| Epitope | Protein | Sequence | % Divergent sites | % Structural divergence |
|---|---|---|---|---|
| 1 | HN | PSKFFTISDSATFAPGPVSNA | 33% | 19% |
| PSKLFIMLDNATFAPGPVVNA | ||||
| 2 | HN | TFRTCFRILALSVQAVTLILVIVTLGELVR | 10% | 63% |
| TFRTCFRILVLSVQAVILILVIVTLGELIR | ||||
| 3 | HN | LSNQLSSI | 0% | N/A |
| 4 | HN | ESATMIASAVGVMNQVIHGVTVSLPL | 8% | 27% |
| ESAAVIASAVGVMNQVIHGVTVSLPL | ||||
| 5 | HN | NQLLATLATICTSQKQVSNCSTNIPLVND | 17% | 7% |
| NQLLSTLATICTNRNQVSNCSTNIPLIND | ||||
| 6 | HN | ATHDFSIGH | 0% | N/A |
| 7 | HN | GCTRIPSFSLKTHWCYTHNVIN | 0% | N/A |
| 8 | HN | SNQYVSMGILVQTA | 14% | 71% |
| SNQYVSMEILAQTA | ||||
| 9 | HN | KTLKIQYLS | 0% | N/A |
| 10 | HN | NRKSCSIATVPDGCAMYCYVST | 0% | N/A |
| 11 | HN | PPTQKLILLFYN | 8% | 58% |
| PPTQKLTLLFYN | ||||
| 12 | HN | WATLVPGVGSG | 9% | 0% |
| WATLVPGAGSG | ||||
| 13 | HN | FENKLIFPAYGGVLPNSTLGVKSAR | 4% | 24% |
| FENKLIFPAYGGVLPNSTLGVKLAR | ||||
| 14 | HN | FFRPVNPYNPCSGP | 0% | N/A |
| 15 | HN | ALRSYFPS | 0% | N/A |
| 16 | HN | FSNRRIQSAFLVCAWNQILVTNCELVVPS | 7% | 3% |
| FSSRRVQSAFLVCAWNQILVTNCELVVPS | ||||
| 17 | HN | EGRVLLINNRLLYYQ | 0% | N/A |
| 18 | HN | WPYELLYEIS | 0% | N/A |
| 19 | HN | SGENVCPTACVSGVYLDPWPLTPYSH | 15% | 38% |
| SGENVCPIVCVSGVYLDPWPLTLYRH | ||||
| 20 | HN | FTGALLN | 0% | N/A |
| 21 | HN | VNPTLYVSALNNLKVLAP | 0% | N/A |
| 22 | HN | GTQGLFAS | 0% | N/A |
| 23 | HN | TTTTCFQ | 0% | N/A |
| 24 | HN | DASVYCVYIM | 0% | N/A |
| 25 | HN | ASNIVGEFQILPVLTR | 6% | 31% |
| ASNIVGEFQILPVLAR | ||||
| SH | SH | MPAIQPPLYLTFLLLILLYLIITLYVWIILTVTYKTSVRHAALYQRSFFHWSFDHSL | 14% | 5% |
| MPAIQPPLYLTFLVLILLYLIITLYVWTILTINYKTSVRYAALYQRSFSRWGFDHSL |
N/A: not applicable.
a Divergence reported in Table 2 reflects only changes in predicted antigenic epitopes. Additional divergent sites are found in non-antigenic regions.
b Top rows contain the consensus sequence across outbreak strains; lower rows contain the JL5 sequence.
Similar outbreaks have occurred in the USA in 2006, 2009–2010, and starting in 2014, continuing through 2017. Outbreak strains are typed and surveilled using the SH sequence, which varies at a higher rate than the rest of the MuV genome (
Orvell et al., 1997
, Yeo et al., 1993
, Takeuchi et al., 1991
). SH is not known to be an antigenic protein, as antibodies against SH are rarely found in the serum of vaccinated or convalescent patients (Ivancic-Jelecki et al., 2008
, Rubin et al., 2012
). As with HN in the Netherlands, the SH amino acid sequences from American outbreak strains indicate that the ongoing outbreak is associated with an emergent clonal lineage belonging to genotype G that is markedly diverse from JL. Previous reports indicate that phylogenetic clustering of strains is consistent whether based on SH, HN, or F, indicating that mutations are largely genotype-specific (Jin et al., 2015
, ). It should therefore be possible to extrapolate findings from HN diversity and potential immune escape to SH. In such a case, sequence diversity or conservation of SH could be considered predictive of diversity in HN, which in turn would have implications for cross-protection or the lack thereof. It is thus potentially prudent to consider that detected variance between outbreak strains and JL in SH is likely mirrored by variance between outbreak strains and JL in the protective epitopes of HN. This argument is bolstered by the current mumps outbreaks in vaccinated individuals in the USA, and the observation that vaccinated mumps patients did not have clear-cut differences in antibody titer from vaccinated, non-patients (Cortese et al., 2011
).Given that at least one American cluster of cases in Arkansas among vaccinated young adults led to an explosion of pediatric cases in a neighboring community with suboptimal vaccine coverage (
Majumder et al., 2017
), the urgency of mumps vaccine efficacy is pressing. As voluntary exemptions from routine vaccinations continue to rise, all rapid and practical measures to ensure full and complete protective immunity among the vaccinated must be considered. Suggestions of a third boost in young adulthood have been made; however, Fiebelkorn et al. demonstrated that the impact of an additional immunization had a negligible long-term effect on antibody titers (Fiebelkorn et al., 2014
). The alternative options include the development of a novel vaccine (Xu et al., 2014
) or formulation of a polyvalent vaccine. The emergence of near-clonal lineages in countries with highly vaccinated populations, as opposed to the diversity of circulating strains in countries such as India where mumps vaccination is optional (Vaidya et al., 2013
), indicates that selection acting on MuV does slowly drive the evolution of new ‘escape variant’ strains. The development of a polyvalent vaccine would be a practical measure that would increase the robustness of vaccine-derived immunity and exponentially decrease the probability of ‘escape variant’-derived outbreaks in vaccinated individuals.Funding
This work was supported by institutional funds from the University of New England.
Ethics statement
This research did not involve the use of human or animal subjects. The authors confirm that all ethical standards for publication have been met.
Conflict of interest
The authors have no conflicts of interest to declare.
Appendix A. Supplementary data
The following is Supplementary data to this article:
References
- Mumps Outbreak at a University and Recommendation for a Third Dose of Measles-Mumps-Rubella Vaccine – Illinois, 2015-2016.MMWR Morb Mortal Wkly Rep. 2016; 65: 731-734https://doi.org/10.15585/mmwr.mm6529a2
- GenBank.Nucleic Acids Res. 2017; 45: D37-D42https://doi.org/10.1093/nar/gkw1070
- Mumps vaccine performance among university students during a mumps outbreak.Clin Infect Dis. 2008; 46: 1172-1180https://doi.org/10.1086/529141
- Mumps antibody levels among students before a mumps outbreak: in search of a correlate of immunity.J Infect Dis. 2011; 204: 1413-1422https://doi.org/10.1093/infdis/jir526
- Localization of a new neutralizing epitope on the mumps virus hemagglutinin-neuraminidase protein.Virus Res. 2001; 74: 133-137
- Persistence of measles, mumps, and rubella antibodies in an MMR-vaccinated cohort: a 20-year follow-up.J Infect Dis. 2008; 197: 950-956https://doi.org/10.1086/528993
- Mumps antibody response in young adults after a third dose of measles-mumps-rubella vaccine.Open Forum Infect Dis. 2014; 1: 094https://doi.org/10.1093/ofid/ofu094
- GOR method for predicting protein secondary structure from amino acid sequence.Methods Enzymol. 1996; 266: 540-553
- Mumps virus F gene and HN gene sequencing as a molecular tool to study mumps virus transmission.Infect Genet Evol. 2016; 45: 145-150https://doi.org/10.1016/j.meegid.2016.08.033
- Mumps-specific cross-neutralization by MMR vaccine-induced antibodies predicts protection against mumps virus infection.Vaccine. 2016; 34: 4166-4171https://doi.org/10.1016/j.vaccine.2016.06.063
- Waning Immunity and Microbial Vaccines — Workshop of the National Institute of Allergy and Infectious Diseases.Clin Vaccine Immunol. 2017; https://doi.org/10.1128/CVI.00034-17
- Are cases of mumps in vaccinated patients attributable to mismatches in both vaccine T-cell and B-cell epitopes?: An immunoinformatic analysis.Hum Vaccin Immunother. 2014; 10: 290-300https://doi.org/10.4161/hv.27139
- Variability of hemagglutinin-neuraminidase and nucleocapsid protein of vaccine and wild-type mumps virus strains.Infect Genet Evol. 2008; 8: 603-613https://doi.org/10.1016/j.meegid.2008.04.007
- Genomic diversity of mumps virus and global distribution of the 12 genotypes.Rev Med Virol. 2015; 25: 85-101https://doi.org/10.1002/rmv.1819
- Trisaccharide containing α2, 3-linked sialic acid is a receptor for mumps virus.Proc Natl Acad Sci U S A. 2016; 113: 11579-11584https://doi.org/10.1073/pnas.1608383113
- Mapping antigenic diversity and strain specificity of mumps virus: a bioinformatics approach.Virology. 2007; 359: 436-446https://doi.org/10.1016/j.virol.2006.09.040
- Persistence of mumps antibodies after 2 doses of measles-mumps-rubella vaccine.J Infect Dis. 2009; 199: 552-560https://doi.org/10.1086/596207
- Vaccine compliance and the 2016 Arkansas mumps outbreak.Lancet Infect Dis. 2017; 17: 361-362https://doi.org/10.1016/S1473-3099(17)30122-6
- Mumps virus neutralizing antibodies do not protect against reinfection with a heterologous mumps virus genotype.Vaccine. 2001; 19: 1727-1731
- Characterization of five conserved genotypes of the mumps virus small hydrophobic (SH) protein gene.J Gen Virol. 1997; 78: 91-95https://doi.org/10.1099/0022-1317-78-1-91
- Mumps Outbreak Among a Highly Vaccinated University Community-New York City, January-April 2014.Clin Infect Dis. 2017; 64: 408-412https://doi.org/10.1093/cid/ciw762
- Molecular biology, pathogenesis and pathology of mumps virus.J Pathol. 2015; 235: 242-252https://doi.org/10.1002/path.4445
- Recent mumps outbreaks in vaccinated populations: no evidence of immune escape.J Virol. 2012; 86: 615-620https://doi.org/10.1128/JVI.06125-11
- Antigenic differences between vaccine and circulating wild-type mumps viruses decreases neutralization capacity of vaccine-induced antibodies.Epidemiol Infect. 2013; 141: 1298-1309
- Identification of conformational neutralization sites on the fusion protein of mumps virus.J Gen Virol. 2015; 96: 982-990https://doi.org/10.1099/vir.0.000059
- Clustal Omega, accurate alignment of very large numbers of sequences.Methods Mol Biol. 2014; 1079: 105-116https://doi.org/10.1007/978-1-62703-646-7_6
- A new mumps virus lineage found in the 1995 mumps outbreak in western Switzerland identified by nucleotide sequence analysis of the SH gene.Arch Virol. 1996; 141: 733-741
- Variations of nucleotide sequences and transcription of the SH gene among mumps virus strains.Virology. 1991; 181: 364-366
- Circulation of two mumps virus genotypes in an unimmunized population in India.J Med Virol. 2013; 85: 1426-1432https://doi.org/10.1002/jmv.23600
- Cross-neutralization between three mumps viruses & mapping of haemagglutinin-neuraminidase (HN) epitopes.Indian J Med Res. 2016; 143: 37-42https://doi.org/10.4103/0971-5916.178587
- Mumps virus nomenclature update: 2012.Weekly Epidemiol Record. 2013; 87: 217-224
- Immunogenicity of novel mumps vaccine candidates generated by genetic modification.J Virol. 2014; 88: 2600-2610https://doi.org/10.1128/JVI.02778-13
- Identification of a new mumps virus lineage by nucleotide sequence analysis of the SH gene of ten different strains.Arch Virol. 1993; 128: 371-377
Article info
Publication history
Published online: October 05, 2017
Accepted:
September 26,
2017
Received in revised form:
September 23,
2017
Received:
August 15,
2017
Corresponding Editor: Eskild Petersen, Aarhus, DenmarkIdentification
Copyright
© 2017 The Author(s). Published by Elsevier Ltd on behalf of International Society for Infectious Diseases.
User license
Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0) | How you can reuse 
Elsevier's open access license policy
Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0)
Permitted
For non-commercial purposes:
- Read, print & download
- Redistribute or republish the final article
- Text & data mine
- Translate the article (private use only, not for distribution)
- Reuse portions or extracts from the article in other works
Not Permitted
- Sell or re-use for commercial purposes
- Distribute translations or adaptations of the article
Elsevier's open access license policy
