Biological characteristics of genetic variants of Urabe AM9 mumps vaccine virus

Virus Research 67 (2000) 49 – 57 www.elsevier.com/locate/virusres

Biological characteristics of genetic variants of Urabe AM9 mumps vaccine virus Kathryn E. Wright *, Kenneth Dimock, Earl G. Brown Department of Biochemistry, Microbiology and Immunology, Faculty of Medicine, Uni6ersity of Ottawa, 451 Smyth Road, Ottawa, Ont., Canada K1H 8M5 Received 11 October 1999; received in revised form 21 January 2000; accepted 21 January 2000

Abstract The Urabe AM9 mumps vaccine is composed of a mixture of variants distinguishable by a difference at nucleotide (nt) 1081 of the hemagglutinin-neuraminidase (HN) gene (Brown, E.G., Dimock, K., Wright, K.E., 1996. The Urabe AM9 mumps vaccine is a mixture of viruses differing at amino acid (aa) 335 of the hemagglutinin-neuraminidase gene with one form associated with disease. J. Infect. Dis. 174, 619 – 622.). Further genetic and biological variation was detected in plaque purified viruses from the Urabe AM9 vaccine by examining the HN gene sequence, plaque morphology, cytopathic effects and growth in Vero cells, and temperature sensitivity (ts). Infection of Vero cells with plaque purified viruses with a G at nt 1081 of the HN gene produced large, clear plaques, caused significant CPE early after infection but yielded lower titres of virus than other purified viruses. None of these viruses were ts. In contrast, half of the plaque purified viruses with an A at nt 1081 were sensitive to a temperature of 39.5°C. These viruses produced small plaques, caused significant CPE and grew to low titres. Two ts viruses possessed a unique aa substitution at aa 468 of HN. The remaining A1081 viruses were not ts, produced large plaques but little CPE, and grew to titres 10-fold higher than the G1081 viruses. Isolates of Urabe AM9 associated with post-vaccination illness were similar to these non-ts A1081 viruses, but could be further sub-divided into two groups on the basis of a difference at aa 464 of HN. The post-vaccination isolates may represent insufficiently attenuated components of the vaccine, while the G1081 and ts subset of A1081 viruses may be more fully attenuated. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Mumps virus; Vaccine strain; HN gene

Mumps virus, like other members of the family Paramyxoviridae, is an enveloped virus with a nonsegmented single stranded RNA genome of negative polarity. Man is the only natural host of * Corresponding author. Tel.: +1-613-5625800; fax: +1613-5625452. E-mail address: [email protected] (K.E. Wright)

mumps virus. Although infection is usually benign, resulting most often in parotitis, the virus spreads systemically and has the capacity to invade visceral organs and the CNS. Asymptomatic involvement of the CNS is common, while the reported incidence of meningitis or meningoencephalitis after infection with mumps can be as high as 17% (Sosin et al., 1989).

0168-1702/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 8 - 1 7 0 2 ( 0 0 ) 0 0 1 2 9 - 5

50

K.E. Wright et al. / Virus Research 67 (2000) 49–57

Several attenuated mumps virus vaccines have been developed in the past 30 years. One of these, Urabe AM9, was introduced in Canada and the UK as part of the Trivirax MMR vaccine (Smith Kline & French). This vaccine was subsequently withdrawn from use because of an unacceptable frequency of post-vaccination meningitis (Forsey et al., 1990; Brown et al., 1991). In order to better understand the insufficient attenuation of the Urabe AM9 virus, comparative sequence analysis of the hemagglutinin-neuraminidase gene (HN) of vaccine and post-vaccination isolates was carried out and it was demonstrated that the Urabe AM9 virus used in the Trivirax vaccine contained at least two viruses distinguishable by a nucleotide (nt) difference at position 1081 of the HN gene (Brown et al., 1996). One type possessed an A (A1081), resulting in a lysine at amino acid (aa) residue 335, which is common to the HN of all mumps viruses for which sequence is available (Yates et al., 1996; Brown and Wright, 1998; Cusi et al., 1998), including the parent Urabe virus (Afzal et al., 1994). Viruses with this sequence were isolated from individuals suffering from post-vaccination illness, indicating a selection in vivo for virus with a lysine at position 335 (Brown et al., 1996; Afzal et al., 1998). The second virus had a G at nt 1081 (G1081), resulting in a glutamic acid at aa 335, and has been isolated from a vaccinee without clinical complications (Afzal et al., 1998). It was speculated that Urabe AM9 virus with this sequence represents an attenuated variant arising during vaccine preparation (reviewed by Brown and Wright, 1998). There are currently no definitive markers for virulence or attenuation of mumps virus, and no model systems for assessing attenuation. In addition, the mumps virus monkey neurovirulence test is not predictive of neurovirulence in humans (Afzal et al., 1999; Rubin et al., 1999). Because Urabe AM9 vaccine appears to contain viruses that differ in the extent of attenuation or virulence, variation in the HN gene and several biological parameters of a panel of Urabe AM9 viruses either obtained from the vaccine by plaque purification, or obtained from cases of post-vaccination disease were examined. We ultimately wanted to understand the extent of genetic varia-

tion of Urabe AM9 viruses with respect to differences in biological properties that correlate with differences in virulence. Previously six viruses were described that had been plaque purified from the Urabe/PM vaccine (Institut Pasteur Merieux, France) in Vero cell monolayers and characterized with respect to the sequence at nt 1081 of HN (Brown et al., 1996). Additional viruses were isolated from the same vaccine by three rounds of plaque purification in Vero cells, to give a total of six A1081 (A3, A5, A9, Aw13, Aw15, Aw1) and six G1081 (G1, G4, G2, G7, Gw4, Gw7) viruses. Stocks of plaque purified viruses used in these experiments underwent no more than five passages in Vero cells, and all genotypes were confirmed in final passage stocks. Viruses were genotyped by MseI digestion of a polymerase chain reaction (PCR) fragment flanking nt 1081 generated with the primers SBL 985PAATTGGGCTACTTTGGT, and SBL 1306-ACTCTTCCTTCTGCACCC, synthesized at the University of Ottawa Biotechnology Institute (Brown et al., 1996). PCR products from viruses with A1081 possessed two cleavage sites, while viruses with G1081 had lost one restriction site, and were cleaved only once, resulting in fragments of distinguishable sizes. Attenuation of viruses, including other paramyxoviruses, is often linked to temperature sensitivity (ts) (Belshe and Hissom, 1982; Crookshanks and Belshe, 1984; Crowe et al., 1994). The ratio of titres of a virus at permissive and nonpermissive temperatures is called the efficiency of plating (EOP). Viruses are considered temperature sensitive when there is at least a 10-fold difference in titres at the two temperatures. The post-vaccination isolates 1004 and 1005, a non-Urabe wildtype virus, CA (CA/73/67, Division of Immunization, Health Canada) and all of the purified A1081 and G1081 viruses were assayed for ts by comparing plaque formation at 37°C, the temperature used for preparation of virus stocks, to plaque formation at 39.5°C using a standard plaque assay (Brown et al., 1996). The G1081 viruses were not ts, while the A1081 viruses varied with respect to ts. CA virus, the post-vaccination Urabe viruses and three of the viruses purified from the vaccine were not ts, while the remaining

K.E. Wright et al. / Virus Research 67 (2000) 49–57

three purified viruses were considered ts (Table 1). Two of these viruses, A9 and A3, had EOP values of B0.02, while titres of the third virus, Aw13, were reduced by approximately 6 – 7-fold at the non-permissive temperature. To rule out the possibility that 37°C was also a non-permissive temperature, these viruses were titrated at 33°C. Titres remained identical to those measured at 37°C, indicating that 37°C is a fully permissive temperature (data not shown). Temperature sensitivity of all purified viruses was initially assessed at the first passage after plaque purification, and confirmed for A9 and A3 after five passages in Vero cells. There are few reports of ts mutants of mumps, and those that have been described arose after passage of virus in rodent cell lines, which are semi-permissive for mumps virus replication Table 1 Efficiency of plating and plaque morphology of mumps viruses. Mumps virusa

Efficiency of plating (titre @ 39.5°C/ titre @ 37°C)

Plaque size (mm) @ 37°C (no. plaques measured)

Urabe AM9 Vac- 1.0 cine Purified G 1081 6iruses G2 1.3 G7 1.2 Gw5 1.9 Gw7 0.7

Mixture

Wild and post-6accination 6iruses CA 0.9 1004 0.8 1005 0.8

Opaque plaques 0.99 0.3 (13) 1.29 0.2 (18) 1.29 0.2 (20)

Purified A 1081 6iruses A5 0.5 Aw1 0.7

Clear plaques 1.490.4 (16) 1.29 0.4 (14) Opaque plaques 1.29 0.3 (23) Small plaques 0.69 0.2 (24) 0.6 90.2 (23) 0.690.1 (26)

Aw15

1.4

Aw13 A3 A9

0.15 0.02 0.003

a

Clear plaques ND 1.390.4 (21) 1.49 0.3 (22) 1.19 0.3 (20)

Serial 10-fold dilutions of each virus were plated on Vero monolayers, and duplicate samples were incubated at 37 and 39.5°C. Plaque production was assessed at day 5 post infection (p.i.). Plaque size was determined after incubation at 37°C, a permissive temperature for all viruses.

51

(Truant and Hallum, 1977; Ogino et al., 1980). The genetic basis of ts of these viruses has not been determined. In general, ts mutations can occur at any site in a viral genome, but changes in the polymerase genes are known to be responsible for the ts phenotype of other ts paramyxoviruses, such as human parainfluenza virus type 3 (Ray et al., 1996; Skiadopoulos et al., 1998) and respiratory syncytial virus (RSV) (Crowe et al., 1994; Firestone et al., 1996). In addition to the differences in growth which have been previously described (Brown et al., 1996), and differences in ts, the purified viruses displayed variation in CPE induction and plaque size in Vero cells which we wished to explore in more detail. The mean plaque diameter from the six plaque purified A1081 viruses, three plaque purified G1081 viruses, two post-vaccination isolates and the wild isolate, CA, were calculated. At day 5 of a plaque assay, Vero cell monolayers were fixed in Carnoy’s solution (25% acetic acid/ 75% methanol, v/v) prior to staining with 0.01% crystal violet and measurement of plaques with calipers. Plaques of three G1081 viruses ranged in size from 1.1 to 1.4 mm in diameter, while within the population of A1081 viruses, there was heterogeneity in plaque size (Table 1). Two viruses, A5 and Aw1, were indistinguishable from G1081 viruses. Aw15 also produced large plaques, but these were opaque rather than clear. The three remaining purified A1081 viruses, A3, A9 and Aw13, which were all ts, had much smaller plaques, with diameters of 0.6 mm, and plaque size remained small at 33°C (data not shown). Viruses from clinical cases of mumps, including post-vaccination Urabe AM9 viruses, had round plaques with diameters of 0.9–1.2 mm, but these were opaque, like those of Aw15, indicating that the plaques contained fewer dead cells than clear plaques. Differences in fusion induced in Vero monolayers after infection with a representative group of viruses were also measured, i.e. purified G1081 and A1081 ts and non-ts, at multiplicity of infection (MOI) 0.05. Fusion was scored at day 3 and 4 post-infection (p.i.), using a modification of a method described by Merz and Wolinsky (1983). Three viruses, G7, Gw7 and a ts virus, A9,

52

K.E. Wright et al. / Virus Research 67 (2000) 49–57

Fig. 1. Fusion in Vero cells after inoculation with Urabe AM9 mumps viruses. Confluent monolayers of Vero cells were inoculated at multiplicity of infection (MOI) = 0.05 with two plaque purified G1081 viruses, Gw7 and G7, a plaque purified A1081 temperature sensitivity (ts) virus, A9, and non-ts virus, Aw15, and two post-vaccination viruses, 1004, 1005. At 72 or 96 h post-infection (p.i.), cells were fixed with Carnoy’s (25% acetic acid, 75% ETOH) and stained with 0.01% crystal violet. Fusion was scored blindly by two individuals, as 0 (absence of cytopathic effects), 1 + (presence of discrete syncytium foci with 4 – 10 nuclei per polykaryote), 2 + (partially confluent syncytium foci containing 10–25 nuclei per polykaryote), 3 + partially confluent syncytium encompassing 50–75% of the field of view), 4 + (confluent syncytium encompassing entire field of view). Values are the means 9 S.D. of values for four to six fields.

showed high fusion scores at day 3 p.i. while the remaining viruses displayed low fusion activity at this time (Fig. 1). By day 4 p.i. this distinction had disappeared, although one A1081 virus, Aw15, still induced very little fusion. It had previously been demonstrated that at day 5 p.i. of Vero cells, post-vaccination and clinical isolates of mumps reached titres 10-fold higher than a stock of vaccine virus, A44, and 5 – 10-fold higher than six viruses purified from the vaccine (Brown et al., 1996). Because increased growth in Vero cells appeared to be a property of post-vaccination Urabe viruses associated with human disease, growth properties were examined in more detail, specifically the kinetics

of replication. Representative viruses from each group, G1081 (Gw7), A1081 ts (A9), A1081 non-ts (A5, Aw15) and a post-vaccination virus (1004) were used to inoculate Vero monolayers at low (0.05) or high (7.0) MOI, and supernatants were collected and titrated at various times p.i. after incubation at 37°C. At low MOI, replication of all the viruses increased after day 1 p.i., peaked and plateaued between days 2 and 4, and declined by day 5 p.i. (Fig. 2A). At all time points except day 5, titres of the G1081 virus were significantly lower than titres of 1004, Aw15 and A5 (P= 0.05). At early time points growth of ts A9 paralleled that of the other A1081 viruses, but growth slowed at days 3 and 4 p.i., when titres remained lower than those for the non-ts A1081 viruses. Growth in Vero cells was assessed at day 4 for the remaining G1081 viruses and a second ts virus, Aw13, and confirmed that all grew to titres B 5× 106/ml after infection at MOI = 0.05. One non-ts A1081virus, Aw1, grew so poorly that it was not possible to prepare stocks with titres higher than 5× 105 pfu/ml, and this virus may represent an additional type of virus present in the vaccine. At high MOI, virus yields were assessed from time 0 to 72 h p.i. (Fig. 2B). The only virus that differed at high MOI relative to low MOI was the ts virus, A9, which grew as well as the other A1081 viruses. The highest titres measured for the A1081 viruses were observed at 48 h p.i., while the highest titre for the G1081 virus was observed at 72 h p.i. Amounts of virus produced were similar to what was observed at low MOI, with the exception of ts A9. Gw7 titres remained significantly lower than those for all the A1081 viruses at 48 and 72 h (P=0.05). Since the original sequence analysis, additional aa substitutions have been described in the HN gene of Urabe viruses purified from a Smith Kline Beecham (Urabe/SKB) vaccine preparation (Afzal et al., 1998). One at nt 1470/aa 464 (Asn“ Lys) was observed in two viruses from cases of postvaccination meningitis, one change was found at nt 1570/aa 498 (Asn “Asp) in a single G1081 virus isolated from a symptomless vaccinee, and the third reported change, nt 343/aa 89 (Met “Val), was associated with small plaque size. The complete cDNA clones of HN were previously se-

K.E. Wright et al. / Virus Research 67 (2000) 49–57

quenced from two post-vaccination Urabe isolates (1004, 1005) and a G1081 virus, and these substitutions had not been observed (Brown et al., 1996). In order to examine the expanded panel of viruses for these changes, PCR fragments derived from the entire HN coding region (with the exception of 21 nt at the 5% end) of two purified viruses, A5 and A9, were sequenced. PCR fragments encompassing nt 1470 and 1570 amplified from isolates associated with post-vaccination meningitis (1004,1005, 719) and post-vaccination parotitis (890, 717) were also sequenced (BB871004, BB871005, BB87890, BB87719, BB87717, Division of Immunization, Health Canada, Ottawa, Canada). The same fragments from two plaque purified G1081 viruses and three other plaque purified A1081 viruses were also sequenced using primers 1258+ : CTGTGCCTGGAATCAGAT and H5%-2: CACGGATCCCACAGGTAGAATTTGGAATTC, (synthesized at the University of Ottawa Biotechnology Institute). Five viruses (one G1081 virus, four A1081 including one post-vaccination isolate) were also examined for the third change at nt 343 (aa 89), using primers

53

Ur+42:CACAATACAACACAGAACCCC and H5%-2. Neither of the two G1081 viruses possessed a change at nt 1570 (Table 2). This is consistent with the findings of Afzal et al. (1998) who reported this change in a G1081 virus isolated from a vaccinee, but not in a G1081 virus plaque purified from Urabe/SKB vaccine. In contrast to the findings of Afzal et al. (1998), no change was observed in any of the viruses at nt 343. The changes which were found fell within a highly variable region of the mumps virus HN gene that extends from aa 461 to aa 477 (Yates et al., 1996; Brown and Wright, 1998). All three viruses isolated from patients with post-vaccination meningitis, 1004, 1005, 719, possessed Lys at aa 464 (Table 2). The stock of 1004, which was not plaque purified, also contained virus with Asn at this residue. The two viruses isolated from patients with post-vaccination parotitis, 890, 717, retained Asn, as did all of the plaque purified A1081 viruses. These results confirm that the Asn“ Lys substitution at aa 464 in the HN gene of viruses possessing A1081 can further differentiate between Urabe viruses associated with post-

Fig. 2. Growth of Urabe AM9 mumps viruses in Vero cells. Triplicate wells of confluent monolayers of Vero cells were inoculated with Urabe AM9 at (A) multiplicity of infection (MOI) = 0.05 or (B) MOI =7.0. At times indicated, supernatants were harvested and assayed for virus in a standard plaque assay. Values represent the mean of three wells 9S.D. Statistical comparisons of means were carried out using a t-test.

54

K.E. Wright et al. / Virus Research 67 (2000) 49–57

Table 2 Urabe AM9 virusesa Group

Isolates

HN sequence nt aa

1. Plaque purified G1081 2. Plaque purified A108 3. Plaque purified A1081 4. Plaque purified A1081 5. Plaque purified A1081 Post-vaccination parotitis 6. Post-vaccination meningitis

G7, Gw7 A3, A9 Aw13 Aw1 A5, Aw15 870,717 1004, 1005, 719

Ts

Plaque size fusion

Growth

1081 335

1470 464

1480 468

G Glu A Lys A Lys A Lys A Lys

C Asn C Asn C Asn nd

G Glu A Lys G Glu Nd

No

Large/+++

B5×106

Yes

Small/+++

B5×106

Yes

Small/ndb

B5×106

No

Large/nd

B106

C Asn

G Glu

No

Large/+

\107

A Lys

A Lys

G Glu

No

Large/+

\107

(+ − ++++)

a Vero cell monolayers were inoculated at multiplicity of infection (MOI) =0.05 with viruses. Total RNA was isolated at 48–72 h post-infection (p.i.) (High Pure Isolation Kit, BM, Laval, Canada) and reversed transcribed with random primers (BM). Polymerase chain reaction (PCR) products extending from nucleotide (nt) 1221–1803 were purified by spin column (GlassMAX DNA System, Life Technologies, Burlington, Canada) and sequenced at the University of Ottawa Biotechnology Institute using a primer annealing to nt 1438–1455; ATATATTCATTCACTCGT. b Not done.

vaccination meningitis and those causing parotitis (Afzal et al., 1998). Loss of the Asn at aa 464 eliminates a potential N-linked glycosylation site, but how this may contribute to virulence is unresolved, as none of the other properties examined discriminated between Urabe viruses associated with parotitis and those associated with meningitis. A second aa change was observed at residue 468 (Glu“ Lys: G“ A at nt 1480) in the HN gene of two of the ts mutants, A3 and A9 (Table 2). Compared to Aw13, which didn’t share this change, these viruses were 10 and 50-fold more sensitive to high temperature. This suggests that the Glu“Lys shift at aa 468 contributes to, but is not uniquely responsible for, ts. This substitution has been reported in two laboratory strains of mumps virus, SBL-1 (Ko¨vamees et al., 1989) and Enders (Yates et al., 1996), and in the Rubini vaccine (Yates et al., 1996). Whether any of these viruses display ts is unknown. None of the changes observed in the HN gene correlated with differences in growth in Vero cells,

plaque size, or fusogenic activity. Increased ability to fuse Vero cells did not correlate with differences in aa at position 335, as both G1081 and A1081 viruses shared this property. Nor did these viruses possess a Ile“Thr substitution at aa 181 of HN, reported to be associated with high fusing variants of RW mumps (Waxham and Wolinsky, 1986; Waxham and Aronowski, 1988). However, high-fusing viruses have been described that retain the Ile at this residue (Ko¨vamees et al., 1989; Takeuchi et al., 1989), and an aa residue in F, aa 195, has been identified as critical for fusion (Tanabayashi et al., 1993). Changes elsewhere in the genome, possibly in the F gene, must be responsible for the differences in fusogenic activity in the Urabe mumps viruses. The mumps virus HN glycoprotein is the major target for neutralizing antibody as measured in vitro (Orvell, 1984). At least four epitopes on HN protein are recognized by neutralizing antibodies (Server et al., 1982; Orvell, 1984). The precise locations of these epitopes on the HN protein are not known, but neutralization escape mutants of

K.E. Wright et al. / Virus Research 67 (2000) 49–57

Kilham and Urabe viruses have been generated with aa substitutions in the region of HN encompassing aa 335 (Ko¨vamees et al., 1990; Yates et al., 1996). More recently, a Urabe specific monoclonal antibody (Mab) was described that neutralizes G1081 but not A1081 viruses (Afzal et al., 1998). Hence, one wished to determine whether there were differences in the immunogenicity of G1081 and A1081 viruses. Polyclonal sera was raised in mice against A9 and G7 viruses, an A1081 nonUrabe clinical isolate KS/67 (Division of Immunization, Health Canada), and a passaged stock of the URA/SKF vaccine. Neutralizing titres (50% inhibition) were calculated for all sera against a wild-type virus (CA), and against homologous and heterologous plaque purified viruses. A pool of three sera specific for G7 had a marginally higher 50% neutralizing titre against a G1081 virus than against the purified A1081 virus, 1/706 compared to 1/610. Similarly, sera raised against A9 virus had a titre of 1/640 against itself, and of 1/513 against the G1081 virus. However, sera raised against these two viruses were able to neutralize the wild isolate as effectively as sera raised against the vaccine; (G7 titre, 1/600; A9 titre, 1/480; vaccine titre, 1/436), indicating that there were no major differences in ability to elicit neutralizing antibody attributable to the described aa substitutions. These results were confirmed by the finding that a panel of HN specific Mabs raised against Kilham or O’Take viruses (Server et al., 1982) bound cells infected with A1081 and G1081 viruses identically in immunofluorescence assays (data not shown). Although Urabe AM9 vaccine was generated by plaque purification (Yamanishi et al., 1973) the results demonstrate that it contains at least five and possibly six unique types of Urabe vaccine virus that can be differentiated by the sequence of the HN gene, plaque morphology, fusion activity, ts and growth potential in Vero cells. Other Urabe variants exist in different preparations of the vaccine, as the small plaque variants purified by Afzal et al. (1998) possessed an aa substitution which was not observed in any of the isolates. It is unlikely that the variants described arose during passage in Vero cells. It had previously been shown that three serial passages of the mixed

55

vaccine in Vero monolayers selects for viruses with G at nt 1081 (Brown et al., 1996), but plaque purified viruses retained their genotype with respect to nt 1081 through at least five passages in Vero cells. As well, ts and growth properties were assessed at early (passage one or two) and late passage (passage five), and were unchanged. The first group of viruses are those with a G at nt 1081 of the HN gene. These viruses, accounting for 25–54% of viruses present in different vaccine preparations (Brown et al., 1996), are not ts, and appear to be homogeneous as assessed by plaque morphology, induction of fusion and growth properties in Vero cells. From the viruses with A at nt 1081, four populations have been plaque purified. One group consists of ts mutants with EOP values of 0.02–0.003 that have a pinpoint plaque morphology, produce early CPE and low growth in Vero cells, at least at low MOI, and possess a unique G“ A substitution at nt 1480 (Glu“ Lys, aa 468). A second group of ts viruses is represented by Aw13, which is less markedly affected by temperatures of 39.5°C (EOP= 0.15), and which does not have the aa substitution at residue 468. The remaining purified A1081 viruses are all non-ts with plaques of the same size as those of G1081 viruses, and may include two groups; viruses that grow so poorly it was not possible to produce stocks for further characterization (Aw1), and viruses that produce delayed CPE but high yields in Vero cells (A5, Aw15). In these properties, this last group of viruses is indistinguishable from the viruses isolated from cases of post-vaccination disease. However, post-vaccination viruses can be further differentiated on the basis of an additional sequence change in HN which correlates with disease severity. Viruses associated with post-vaccination parotitis are identical in HN sequence to the purified non-ts A1081 viruses, while those associated with post-vaccination meningitis may represent a sixth population of A1081 viruses with a C“A substitution at nt 1470 (Asn “ Lys, aa 464). This virus has yet to be purified from the vaccine, so it cannot be determined whether it is a rare variant that exists in the vaccine and is selected during human infection or whether it is generated during infection.

56

K.E. Wright et al. / Virus Research 67 (2000) 49–57

What is the relevance of the observed biological differences in these various virus isolates? Given the lack of markers for virulence or attenuation of mumps viruses, it is difficult to answer this question. The results indicate that the Urabe vaccine is a heterogeneous mixture of G1081 and A1081 viruses with varying extents of attenuation. Virulent Urabe viruses, all A1081, display delayed fusogenicity and high growth in Vero cell monolayers. Viruses possessing G1081 are assumed to be attenuated as they have never been associated with post-vaccination disease (Brown et al., 1991) but have been isolated from an individual without complications after vaccination (Afzal et al., 1998). It was also predicted that the ts viruses are attenuated, based on the fact that such strains have not been isolated with post-vaccination viruses. Association of ts with attenuation has been observed for other paramyxoviruses, including HPIV3 (Crookshanks and Belshe, 1984; Hall et al., 1992) and RSV (Crowe et al., 1994) where changes in the polymerase gene appear to be responsible for sensitivity to high temperature. However, lesions in other genes also contribute to attenuation (Hall et al., 1992). As well as being attenuated, both the G1081 virus and a ts A1081 virus are as immunogenic as the vaccine and wild-type virus, at least in mice, and so both would be good vaccine candidates. A third population of attenuated viruses may be represented by the poorly replicating non-ts A1081 virus Aw1. Although the sequence changes described in HN are reliable markers for viruses with different biological properties in vitro, further studies are needed to assess their roles in these biological processes. Future work will focus on understanding the importance of the genetic variation which has been described in the HN gene, if any, in effecting the biological properties of ts, plaque morphology and growth potential of Urabe viruses.

Acknowledgements The authors would like to thank Dr J. Wolinsky for the gift of anti-HN monoclonal antibodies. This work was funded in part by funds

received from the National Sciences and Engineering Research Council of Canada. The authors hold a US patent covering the G1081 form of Urabe AM9 virus.

References Afzal, M.A., Pickford, A.R., Yates, P.J., Forsey, T., Minor, P.D., 1994. Matrix protein gene sequence of vaccine and vaccine-associated strains of mumps virus. J. Gen. Virol. 75, 1169 – 1172. Afzal, M.A., Yates, P.J., Minor, P.D., 1998. Nucleotide sequence at position 1081 of the hemagglutinin-neuraminidase gene in the mumps Urabe vaccine strain. J. Infect. Dis. 177, 265 – 266. Afzal, M.A., Marsden, S.A., Hull, R.M., Pipkin, P.A., Bentley, M.L., Minor, P.L., 1999. Evaluation of the neurovirulence test for mumps vaccines. Biologicals 27, 43 – 49. Belshe, R.B., Hissom, F.K., 1982. Cold adaptation of parainfluenza virus type 3: induction of three phenotypic markers. J. Med. Virol. 10, 235 – 242. Brown, E.G., Wright, K.E., 1998. Genetic studies on a mumps vaccine strain associated with meningitis. Rev. Med. Virol. 8, 129 – 142. Brown, E.G., Furesz, J., Dimock, K., Yarosh, W., Contreras, G., 1991. Nucleotide sequence analysis of the Urabe mumps vaccine strain that caused meningitis in vaccine recipients. Vaccine 9, 840 – 842. 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. Infect. Dis. 174, 619 – 622. 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. Crowe, J.E. Jr, Bui, P.T., Davis, A.R., Chanock, R.M., Murphy, B.R., 1994. A further attenuated derivative of a cold-passaged temperature sensitive mutant of human respiratory syncytial virus retains immunogenicity and protective efficacy against wild-type challenge in seronegative chimpanzees. Vaccine 12, 783 – 790. Cusi, M.G., Santini, L., Bianchi, S., Valassina, M., Valensin, P.E., 1998. Nucleotide sequence at position 1081 of the hemagglutinin-neuraminidase gene in wild-type strains of mumps virus is the most relevant marker of virulence. J. Clin. Microbiol. 36, 3743 – 3744. Firestone, C.-Y., Whiteheade, S.S., Collins, P.L., Murphy, B.R., Crowe, J.E. Jr, 1996. Nucleotide sequence analysis of the respiratory syncytial virus subgroup A cold-passaged (cp) temperature sensitive (ts) cpts-248/404 live attenuated virus vaccine candidate. Virology 225, 419 – 422. Forsey, T., Mawn, J.A., Yates, P.J., Bentley, M.L., Minor, P.D., 1990. Differentiation of vaccine and wild mumps

K.E. Wright et al. / Virus Research 67 (2000) 49–57 viruses using the polymerase chain reaction and dideoxynucleotide sequencing. J. Gen. Virol. 71, 987–990. Hall, S.L., Stokes, A., Tierney, E.L., London, W.T., Belshe, R.B., Newman, C., Murphy, B.R., 1992. Cold-passaged human parainfluenza type 3 viruses contain ts and non-ts mutations leading to attenuation in rhesus monkeys. Virus Res. 22, 173 – 184. Ko¨vamees, J., Norrby, E., Elango, N., 1989. Complete nucleotide sequence of the hemagglutinin-neuraminidase (HN) mRNA of mumps virus and comparison of paramyxovirus HN proteins. Virus Res. 12, 87–96. Ko¨vamees, J., Rydbeck, R., Orvell, C., Norrby, E., 1990. Hemagglutinin-neuraminidase (HN) amino acid alterations in neutralization escape mutants of Kilham mumps virus. Virus Res. 17, 119 – 130. Merz, D.C., Wolinsky, J.S., 1983. Conversion of nonfusing mumps virus infections to fusing infections by selective proteolysis of the HN glycoprotein. Virology 131, 328– 340. Ogino, T., Kotake, S., Yoshida, T.O., 1980. Persistent infection of mouse tumor cells with mumps virus. Biken J. 23, 15– 24. Orvell, C., 1984. The reactions of monoclonal antibodies with structural proteins of mumps virus. J. Immunol. 132, 2622 – 2629. Ray, R., Galinski, M.S., Neminway, B.R., Meyer, K., Newman, F.K., Belshe, R.B., 1996. Temperature-sensitive phenotype of the human parainfluenza virus type 3 candidate vaccine strain (cp45) correlates with a defect in the L gene. J. Virol. 70, 580 – 584. Rubin, S.A., Snoy, P.J., Wright, K.E., Brown, E.G., Reeve, P., Beeler, J., Carbone, K.M., 1999. The mumps neurovirulence safety test in rhesus monkeys: a comparison of mumps virus strains. J. Infect. Dis. 180, 521–525. Server, A.C., Merz, D.C., Waxham, M.N., Wolinsky, J.S., 1982. Differentiation of mumps virus strains with monoclonal antibody to the HN glycoprotein. Infect. Immun. 35, 179 – 186.

57

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 2 cp45 live attenuated vaccine candidate contribute to its temperature sensitive and attenuation phenotypes. J. Virol. 72, 1762 – 1768. Sosin, D.M., Cochi, S.L., Gunn, R.A., Jennings, C.E., Preblud, S.R., 1989. Changing epidemiology of mumps and its impact on university campuses. Pediatrics 84, 779 – 784. Takeuchi, K., Tahabayashi, K., Hishiyama, M., Yamada, A., Sugiura, A., 1989. Cloning and sequencing of the hemagglutinin-neuraminidase gene of mumps virus (Miyahara strain). Nucleic Acid Res. 17, 5840. Tanabayashi, K., Takeuchi, K., Okasaki, K., Hishiyama, M., Yamada, A., 1993. Identification of an amino acid that defines the fusogenicity of mumps virus. J. Virol. 67, 2928 – 2931. Truant, A.L., Hallum, J.V., 1977. A persistent infection of baby hamster kidney-21 cells with mumps virus and the role of temperature-sensitive variants. J. Med. Virol. 1, 49 – 67. Waxham, M.N., Aronowski, J., 1988. Identification of amino acids involved in the sialidase activity of the mumps hemagglutinin-neuraminidase protein. Virology 167, 226 – 232. Waxham, M.N., Wolinsky, J.S., 1986. A fusing mumps virus variant selected from a nonfusing parent with the neuraminidase inhibitor 2-deoxy-2,3-dehydro-N-acetylneuraminic acid. Virology 151, 286 – 295. Yamanishi, K., Takahashi, M., Ueda, S., Minekawa, Y., Ogino, T., Suzuki, N., Okuno, Y., 1973. Studies on live mumps virus vaccine. V. Development of a new mumps vaccine ‘AM9’ by plaque cloning. Biken J. 16, 161 – 166. Yates, P.J., Afzal, M.A., Minor, P.D., 1996. Antigenic and genetic variation of the HN proteins of mumps virus strains. J. Gen. Virol. 77, 2491 – 2497.

.