- Email: [email protected]
Foetal cross-protection experiments between type 1 and type 2 bovine viral diarrhoea virus in pregnant ewes
Veterinary Microbiology 64 (1999) 185±196
Foetal cross-protection experiments between type 1 and type 2 bovine viral diarrhoea virus in pregnant ewes D.J. Paton*, G. Sharp, G. Ibata Central Veterinary Laboratory, Veterinary Laboratories Agency, Addlestone, Surrey KT15 3NB, UK
Abstract A flock of 82 non-pregnant ewes was split into three immunisation groups and given an intranasal dose of either cell culture medium, or a type 1 or a type 2 bovine viral diarrhoea virus (BVDV-1 or BVDV-2). Two months later the flock was reconstituted and after a further three weeks, the ewes were bred to pestivirus negative rams after synchronisation of oestrus using progesterone sponges. Fifty-five ewes were segregated into three challenge groups, each of which comprised ewes from different immunisation groups. At 7 weeks gestation, one challenge group was given an intranasal dose of cell culture medium, whilst the other two were given intranasal doses of either BVDV-1 or BVDV-2, using the same inocula as for the immunisations. Three weeks later, the ewes were killed and their foetuses tested for the presence of BVDV-1 and BVDV-2. The results showed that immunisation of six ewes without subsequent challenge did not lead to infection of any of their 11 foetuses. Challenge with BVDV-1 or BVDV-2 in the absence of immunisation lead to 15 out of 15 or 11 out of 14 foetuses becoming infected, respectively. Immunisation with the homologous virus to that used for challenge resulted in complete protection of 32 foetuses from 15 ewes. Heterologous protection was one way. All 12 foetuses from ewes immunised with BVDV-1 were protected from challenge with BVDV-2, whereas 18 foetuses from ewes immunised with BVDV-2 were all infected after challenge with BVDV-1. This provides evidence that a recent exposure to infection with one pestivirus does not necessarily induce foetal protection against another. The one-way result suggests that factors other than antigenic differences are involved in cross-protection. Crown Copyright # 1999 Published by Elsevier Science B.V. Keywords: Pestivirus; BVD; Sheep model; Foetal infection; Cross-protection
* Corresponding author. Tel.: +44-1932-357285; fax: +44-1932-357239; e-mail: [email protected] 0378-1135/99/$ ± see front matter Crown Copyright # 1999 Published by Elsevier Science B.V. PII S 0 3 7 8 - 1 1 3 5 ( 9 8 ) 0 0 2 6 9 - 7
186
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
1. Introduction Bovine viral diarrhoea virus (BVDV) is a small, enveloped, single stranded, positive sense RNA virus of the Pestivirus genus within the family Flaviviridae. Two types of BVDV (BVDV-1 and BVDV-2, synonym Pestivirus 1 and Pestivirus 2) can be distinguished from one another and from the two other principal pestivirus types, namely ovine border disease virus (BDV or Pestivirus 3) and classical swine fever virus (CSFV or Pestivirus 4). All of the pestiviruses are serologically related and many can infect a variety of artiodactyl hosts. Infection of cattle with BVDV is common in most parts of the world. Acute infection is characterised by a period of systemic viral infection and virus shedding, which is followed by recovery and development of long lasting antibody. Although acute infection is frequently unnoticed, it can cause or contribute to a variety of problems, including diseases of the reproductive, respiratory, and enteric systems (Baker, 1995). In cases of so-called virulent BVDV, there may be haemorrhagic disease and significant mortality in both young and adult animals. BVDV readily crosses the placenta and can cause infertility and embryonic or foetal death. Foetuses infected before the onset of immunocompetence can become persistently infected. Persistently infected (PI) animals may have reduced viability both in utero and after birth, but a proportion of PI animals survive, acting as an important reservoir of infection for other cattle. Prevention of foetal infection is, therefore, an important aim in BVD control and relies on either avoiding exposure of pregnant cattle to viral challenge, or immunisation before pregnancy. Many cattle develop an immunity to BVDV, either as a consequence of an earlier acute infection, or because of vaccination with available killed or live vaccines. It is known that prior infection or vaccination of the dam can protect foetuses from challenge in pregnancy with homologous (Duffell et al., 1984) or even somewhat dissimilar strains of BVDV (Brownlie et al., 1995), but there is also evidence that this protection may be incomplete (Harkness et al., 1987; Meyling et al., 1987). The antigenic and genetic diversity of BVDV is increasingly recognised. BVDV-1 has been the predominant type in most of the world, whereas BVDV-2 appears to be a newly emerging variant, so far only common in North America. Both types of BVDV can be associated with acute and persistent infections that range from being subclinical to fatal. However, in North America, the emergence of BVDV-2 has been associated with outbreaks of severe and sometimes haemorrhagic disease characterised by fatal acute infections (Pellerin et al., 1994; Ridpath et al., 1994). In Britain, the predominant pestivirus isolated from cattle has been BVDV-1 (Paton et al., 1995). A large proportion of adult cattle are seropositive due to earlier exposure to the virus (Harkness et al., 1978), and a killed vaccine based on BVDV-1 is starting to be used (Brownlie et al., 1995). However, serological cross-neutralisation between BVDV-1 and BVDV-2 is incomplete, both in vitro (Pellerin et al., 1994; Paton et al., 1995; Dekker et al., 1995) and in vivo (Bolin and Ridpath, 1995). There is, therefore, considerable interest to know whether the widespread immunity to BVDV-1 will protect pregnant cattle against challenge with BVDV-2. In the British sheep industry few measures are taken to control pestivirus infections. Characterisation of isolates has shown that two species of pestivirus predominate, namely
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
187
BVDV-1 and BDV, but that BVDV-2 has also been isolated (Paton et al., 1994; Vilcek et al., 1997). Previous studies have used sheep as a model to test the efficacy of BVDV immunisation and cross-protection (Vantsis et al., 1980; Carlsson et al., 1991; Bruschke et al., 1996). The aim of this study was to use sheep as a model to examine foetal crossprotection between infection with BVDV-1 and BVDV-2. 2. Materials and methods 2.1. Viruses Two non-cytopathic BVD viruses originally isolated from PI cattle were used for immunisation and challenge of sheep, and for in vitro serology. They were selected on the basis of their known antigenic and genetic classification (Paton, 1995; van Rijn et al., 1997), and for their proven ability to infect ovine foetuses (Bruschke et al., 1996). Virus Pest-1/Bo/ID-DLO-4800/NL-1993, a BVDV-1b isolated in The Netherlands in 1993 was kindly donated by Christianne Bruschke as a cell culture supernatant at passage one. It was passaged once more in bovine turbinate (BT) cells to give a stock with a titre of 105.6TCID50/ml. Pest-2/Bo/CVL-178003/GB-1987 is a BVDV-2 isolated in England in 1987. The stock used had been passaged three times in BT cells and had a titre of 107.8 TCID50/ml. The same inocula were used for both immunisation and challenge, each ewe receiving 2 ml of virus diluted to 105 TCID50/ml in MEM 10% FCS. 2.2. Experimental design A flock of 82 cross-bred ewes were used for the study. All were blood tested for pestivirus status using serology and attempted virus isolation prior to use. They were maintained on 20 ha of pasture throughout the experiment. Groups were segregated by placing them in securely fenced fields of approximately 4 ha, with an empty field between each group. Separate protective clothing was used for each group and animal handling and blood tests for each group were carried out on different days. The segregation of the flock into groups for immunisation and challenge is summarised in Fig. 1. In September, the flock was split randomly into three immunisation groups NI (no immunisation), IV1 (immunisation with BVDV-1) and IV2 (immunisation with BVDV-2). Each group comprised 26 principal ewes, but two additional ewes were left unimmunised in each of groups IV1 and IV2 in order to examine the efficiency of horizontal virus spread from the immunised animals. Immunisation consisted of an intranasal dose of either cell culture medium (MEM plus 10% FCS), BVDV-1 or BVDV2. A luer-lock syringe fitted with a spray nozzle was used to apply 1 ml of inoculum into each nostril. Blood samples were taken for serology approximately 3 and 6 weeks after immunisation. Six weeks after immunisation, the two unimmunised ewes within groups IV1 and IV2 were removed and housed in isolation elsewhere. At 8 weeks after immunisation the flock was reconstituted as a single group, and after a further 3 weeks, the ewes were bred to four rams after synchronisation of oestrus using
188
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
Fig. 1. Summary of cross-protection study design to show segregation of ewes into immunisation and challenge groups.
progesterone sponges. For this purpose, the ewes were split into five groups, each group being served by the rams during a single 24 h period on 5 consecutive days. One week later the ewes were divided into two groups, each of which was left with two rams for 3 weeks. Blood samples for serology and virology were taken from the rams immediately before their first introduction. At 5 weeks after breeding, ultrasound testing was used to select 55 ewes which were all at approximately the same stage of gestation. These ewes were segregated into three challenge groups: NC (no challenge), CV1 (challenge with BVDV-1) and CV2 (challenge with BVDV-2). The NC group comprised three ewes from each immunisation group. The CV1 and CV2 groups each comprised eight ewes from NI, eight ewes from IV1 and seven ewes from IV2 (Fig. 1). Ewes were assigned to challenge groups so as to match the range of antibody titres present in animals within each challenge group. At 7 weeks gestation, and 120 days after the initial immunisations, one challenge group was given an intranasal dose of cell culture medium, whilst the other two were given intranasal doses of either BVDV-1 or BVDV-2, using the same inocula and procedure as for the immunisations. Three weeks later, the ewes were killed and their foetuses tested for the presence of BVDV-1 and BVDV-2. For this purpose, an aseptic technique was used to collect umbilical blood, and a portion of spleen and kidney from each foetus. Foetal crown-rump lengths were recorded as an indicator of foetal age.
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
189
2.3. Test procedures A microtitre plate format of the neutralising peroxidase linked assay (NPLA) was used for serology (Hyera et al., 1987). Doubling dilutions of sera, starting at 1 in 10 were incubated for 1 h at 378C with equal volumes of medium containing approximately 100 TCID50 of either Pest-1/Bo/ID-DLO-4800/NL-1993 or Pest-2/Bo/CVL-178003/GB1987. BT cells were added and after 5 days further incubation, the presence of virus was tested for by peroxidase linked assay (PLA Holm Jensen, 1981) using a hyperimmune, bovine BVDV antiserum and a peroxidase linked rabbit anti-bovine conjugate (Dako). Titres are expressed as reciprocal dilutions of sera that neutralise 50% of replicate virus-infected monolayers. For calculation of geometric means, antibody titres were converted to logarithmic values and for this purpose, titres of <10 were given an arbitrary log value of 0.5. Virus isolation was attempted from pre-inoculation ewe sera, pre-breeding ram sera, and from post mortem collections of foetal and maternal sera, foetal spleen and kidney. Cell cultures and gamma irradiated foetal calf serum used for their growth had been previously screened and shown to be free of adventitious BVDV and BVDV inhibitors. Fresh serum samples in 10 ml aliquots were each applied directly to four wells of a microtitre plate, and then incubated with BT cells for 5 days at 378C. The presence of virus was detected by PLA, as described above. For each foetus, a pool of freshly collected spleen and kidney was homogenised and inoculated into flasks containing monolayers of rapidly growing BT cells. After 5 days incubation, the cultures were subpassaged into microtitre plate wells, incubated another 5 days and then fixed and immunostained for virus growth and detection as described for sera. In the case of isolation from tissues, additional immunostaining was carried out with monoclonal antibodies able to distinguish between the BVDV-1 and BVDV-2 viruses in the PLA. The monoclonal antibodies WB160 and WB166 are selective for Pest-1/Bo/ID-DLO-4800/ NL-1993, whilst WB437 and WB537 are selective for Pest-2/Bo/CVL-178003/GB-1987 (Paton et al., 1995). 2.4. Statistics The significance of differences in pregnancy rates and foetal infection rates between groups was analysed statistically by Fisher exact test. 3. Results 3.1. Response to immunisation All ewes were pestivirus negative and seronegative to both BVD viruses (titre of <10) when tested at 2 months prior to immunisation. All were seronegative at the time of immunisation. Ewes in the control NI group were still seronegative at the time of challenge, and those that received neither immunisation nor challenge remained seronegative throughout the experiment. All of the immunised ewes seroconverted
190
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
Table 1 Results of antibody tests and foetal virus tests on unchallengeda ewes from different immunisation groups Ewe
Immunisation
Antibody at challengea
Antibody at necropsy
BVDV-1
BVDV-1
BVDV-2
Foetuses infected
BVDV-2
H136 F127 F83
None (NI) None (NI) None (NI)
<10 <10 <10
<10 <10 <10
<10 <10 <10
<10 <10 <10
0/2 0/3 0/2
Mean
None (NI)
3
3
3
3
0%
G61 G289 F223
BVDV-1 (IV1) BVDV-1 (IV1) BVDV-1 (IV1)
1280 320 960
<10 <10 <10
1280 2560 1280
20 <10 <10
0/2 0/2 0/2
Mean
BVDV-1 (IV1)
741
3
1620
6
0%
E698 F256 G675
BVDV-2 (IV2) BVDV-2 (IV2) BVDV-2 (IV2)
10 <10 <10
80 320 480
<10 <10 <10
60 640 640
0/1 0/2 0/2
Mean
BVDV-2 (IV2)
5
229
3
293
0%
a
Antibody at challenge at time of challenge of other groups i.e. 120 days after immunisation.
within 6 weeks, and levels of neutralising antibody had reached a plateau by the time of challenge. The two uninoculated ewes within each of groups IV1 and IV2 had not seroconverted by the time that they were removed from the other ewes 6 weeks after the immunisation procedure. They were still seronegative 4 months later. Individual clinical examinations were not performed, but no gross signs of illness were observed in the ewes after immunisation. A single ewe from the IV2 group died of pneumonia 11 weeks after immunisation. Antibody levels at the time of challenge are shown in Tables 1±3. In general, higher titres were observed in ewes inoculated with BVDV-1 than in ewes inoculated with BVDV-2. The homologous, geometric mean antibody titre for all animals immunised with BVDV-1 was 1043. For ewes given BVDV-2, the equivalent homologous mean titre was 172. Antibody titres were much higher to the homologous virus than to the heterologous virus. Mean heterologous titres in the BVDV-1 and BVDV-2 immunised animals at the time of challenge were 6 and 5, respectively. 3.2. Ram testing and pregnancy diagnosis The four rams were pestivirus negative and seronegative when introduced to the first ewes. Ultrasound testing at approximately 5 weeks after breeding confirmed that 58 ewes were at the same stage of gestation. The proportions of ewes at this stage of pregnancy were not statistically different between immunisation groups: 20 out of 26, 17 out of 26 and 21 out of 25 for groups NI, IV1 and IV2, respectively.
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
191
Table 2 Results of antibody tests and foetal virus tests on ewes from different immunisation groups and challenged with BVDV-1 Ewe
Immunisation
(NI) (NI) (NI) (NI) (NI) (NI) (NI) (NI)
Antibody at challenge
Antibody at necropsy
BVDV-1
BVDV-1
<10 <10 <10 <10 <10 <10 <10 <10
BVDV-2
C161 D63 E619 F102 F492 G220 G510 G672
None None None None None None None None
<10 <10 <10 <10 <10 <10 <10 <10
Mean
None (NI)
3
3
D636 E253 E693 E757 F85 F221 G658
BVDV-1 BVDV-1 BVDV-1 BVDV-1 BVDV-1 BVDV-1 BVDV-1
(IV1) (IV1) (IV1) (IV1) (IV1) (IV1) (IV1)
2560 960 640 960 480 2560 2560
<10 <10 <10 20 <10 <10 <10
Mean
BVDV-1 (IV1)
1259
4
D631 E262 F124 F67 F91 G336 G357 G50
BVDV-2 BVDV-2 BVDV-2 BVDV-2 BVDV-2 BVDV-2 BVDV-2 BVDV-2
(IV2) (IV2) (IV2) (IV2) (IV2) (IV2) (IV2) (IV2)
<10 <10 60 <10 15 <10 <10 <10
Mean
BVDV-2 (IV2)
6
640 240 2560 320 120 60 240 960
Foetuses infected
BVDV-2 <10 <10 10 <10 <10 <10 <10 <10
2/2 1/1 2/2 2/2 3/3 2/2 2/2 1/1
355
4
100%
1280 640 960 1280 480 3840 2560
10 15 15 30 30 <10 15
0/2 0/3 0/2 0/2 0/2 0/2 0/4
1259
13
0%
120 160 480 160 480 120 40 240
3840 3840 15360 2560 10240 7680 3840 2560
5120 3840 1280 2560 20480 15360 2560 2560
3/3 2/2 2/2 2/2 2/2 3/3 2/2 2/2
178
5158
4340
100%
3.3. Effect of challenge No illness was observed in ewes after challenge. At post-mortem, all 55 ewes selected for the challenge study were found to be pregnant with one to three foetuses ranging in size from 12.5 to 18 cm of crown rump length. Gross foetal abnormalities were not apparent. Sera collected from ewes post-mortem (3 weeks after challenge) were all virus negative apart from one sample. This ewe (F125 ± Table 3) had been challenged with BVDV-2 without prior immunisation. A summary of the results of attempted virus isolation from foetal serum and organs is shown in Tables 1±3. Monoclonal antibody typing demonstrated that recovered virus was of the type used for challenge rather than immunisation. Eighteen foetuses obtained from nine ewes that had not been challenged in pregnancy were all virus negative, regardless of immunisation status. All 15 foetuses from eight unimmunised ewes that had been challenged with BVDV-1 were found to be infected, virus being isolated from both serum
192
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
Table 3 Results of antibody tests and foetal virus tests on ewes from different immunisation groups and challenged with BVDV-2 Ewe
Immunisation
(NI) (NI) (NI) (NI) (NI) (NI) (NI) (NI)
Antibody at challenge
Antibody at necropsy
BVDV-1
BVDV-1
<10 <10 <10 <10 <10 <10 <10 <10
BVDV-2
H100 D629 E421 F75 F125 G52 G487 G663
None None None None None None None None
<10 <10 <10 <10 <10 <10 <10 <10
Mean
None (NI)
3
3
E587 H440 F510 J321 F507 G333 G331
BVDV-1 BVDV-1 BVDV-1 BVDV-1 BVDV-1 BVDV-1 BVDV-1
(IV1) (IV1) (IV1) (IV1) (IV1) (IV1) (IV1)
960 7680 960 480 640 480 480
<10 20 <10 15 40 <10 20
Mean
BVDV-1 (IV1)
936
10
F111 E264 E686 F77 H776 F72 H541 H219
BVDV-2 BVDV-2 BVDV-2 BVDV-2 BVDV-2 BVDV-2 BVDV-2 BVDV-2
(IV2) (IV2) (IV2) (IV2) (IV2) (IV2) (IV2) (IV2)
20 <10 <10 <10 <10 <10 <10 <10
Mean
BVDV-2 (IV2)
4
<10 <10 <10 <10 <10 10 <10 15
Foetuses infected
BVDV-2 <10 <10 <10 40 <10 <10 <10 <10
0/1 0/1 1/1 2/2 2/2 2/2 2/2 2/3
4
4
79%
640 2560 640 480 640 320 320
<10 40 15 <10 20 <10 10
0/1 0/2 0/2 0/2 0/1 0/2 0/2
611
9
0%
640 80 120 960 1280 120 <10 60
10 <10 <10 <10 <10 <10 <10 <10
640 240 160 960 640 160 10 80
0/2 0/2 0/2 0/2 0/2 0/2 0/1 0/2
145
4
194
0%
and tissues in every case. A lesser proportion (11 out of 14) of foetuses were infected in the unimmunised ewes challenged with BVDV-2, and in some cases virus was recovered only from serum or only from tissues. This difference is not statistically significant, either on a ewe basis (p 0.47) or on a foetus basis (p 0.1). None of the foetuses were infected from ewes that had been challenged with the same virus as was used for prior immunisation. Likewise, none of the foetuses were infected from ewes previously immunised with BVDV-1 and then challenged with BVDV-2. However, all of the foetuses were infected from ewes that were immunised with BVDV-2 and then challenged with BVDV-1, and this difference is statistically significant. Serological results for samples collected at post mortem are shown in Tables 1±3. Unimmunised ewes challenged with BVDV-1 had seroconverted strongly when killed 3 weeks later (titre range 60±2560, Table 2), whereas unimmunised ewes that received BVDV-2 were mostly still seronegative at death (titre range <10±40, Table 3). Of the four pre-immunised groups that were challenged, either with a homologous or a heterologous
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
193
virus, a significant titre increase was only seen between challenge and death in one group. This was the group immunised with BVDV-2 and challenged with BVDV-1 (Table 2). At the final collection, these ewes had very high antibody titres to both BVDV1 and BVDV-2. 4. Discussion Use of sheep rather than cattle is much less costly for the in vivo study of crossprotection between pestiviruses, especially for work involving pregnant animals. Furthermore the predominant pestivirus types found in cattle also occur naturally in sheep, whereas the reverse is not true, since BDV is very rarely if ever isolated from cattle. A previous study had shown that the two bovine isolates Pest-1/Bo/ID-DLO-4800/ NL-1993 and Pest-2/Bo/CVL-178003/GB-1987 were able to infect ovine foetuses following intranasal exposure of their dams in early pregnancy (Bruschke et al., 1996). In the present experiment, a similar approach was used, except that sheep were maintained in groups at pasture rather than indoors in isolation units. This reduces costs further and may be less stressful for the animals. The results indicated that seronegative sheep kept in separate pastures on the same farm did not become infected from other inoculated groups. Indeed, horizontal transmission of virus did not occur from inoculated sheep to four control ewes kept in the same fields. Other reports have documented inefficient spread of ruminant pestiviruses between acutely infected ruminants and pastured sheep (Nettleton, 1987; McGowan et al., 1993). Intranasal inoculation of BVDV is a practical means to mimic the natural route of infection, whilst allowing for a controlled exposure to virus (McGowan et al., 1993). Selection of an appropriate dose properly requires prior titration of the inoculum in the target species (Richardson et al., 1976) and establishment of the normal level of field exposure. The dose chosen in the present study was the same as previously used by Bruschke et al. (1996) and reflects the level of virus shed from the nose of PI animals ± commonly in the region of 104±105 TCID50/ml. The BVDV-1 inoculum appeared to be more efficient than the BVDV-2 inoculum at infecting the foetuses of seronegative ewes (100% vs. 79% for the BVDV-2), although due to the relatively small group sizes the difference is not statistically significant. It also induced earlier seroconversion with higher titres of neutralising antibodies. Differences in immunogenicity may amongst other factors depend upon different patterns of viral replication, and this could also account for differences in foetal penetration. The majority of acute pestivirus infections in ruminants are mild or subclinical and little is known concerning differences in the replication or pathogenicity of the viruses involved. Isolates of extreme virulence or attenuation have been documented (Lobmann et al., 1986; Corapi et al., 1989), and subtler differences, for instance in the replication of bovine and ovine isolates in lambs, have also been reported (Jewett et al., 1990). Both of the viruses used in our inocula had been passaged only a small number of times in cell culture to minimise problems of attenuation. Both were also diluted so as to give the same titre of infectivity, as measured in bovine turbinate cell cultures, and titrations performed in ovine testicle cells gave similar results (data not shown). It would nevertheless be interesting to test higher doses
194
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
of the BVDV-2 inoculum in sheep, to see if this resulted in more extensive replication and a stronger seroconversion. Richardson et al. (1976) demonstrated a dose-dependent severity of congenital disease in lambs due to experimental pestivirus infection. With respect to immunogenicity, it has been shown that non-cytopathic biotypes of BVDV induce a stronger immunity than cytopathic viruses (Lambot et al., 1997), but otherwise there is not much information available on strain-specific differences. Jewett et al. (1990) described differences in the neutralising antibody responses of lambs to different pestivirus isolates, all given at the same dose. However, it was not clear how much these differences were due to pathogenic variability amongst the challenge viruses or to variable antigenic mismatch between the challenge viruses and the virus used for the neutralisation test. As expected, both groups of ewes that had been pre-immunised, did not transmit virus to their foetuses after a homologous virus challenge. Similar results have been reported for sheep, using BVDV and BDV (Vantsis et al., 1980) and for cattle using BVDV (Duffell et al., 1984), but this is the first such report concerning BVDV-2. However, in these experiments, the period between immunisation and challenge has been relatively short, and the duration of protective immunity has not been established. Our homologous challenge experiments give no indication of a particular titre of neutralising antibody that would equate to foetal protection. All of our BVDV-1 immunised sheep had moderately high titres of homologous neutralising antibody at the time of challenge (titres of 480), but the equivalent levels of antibody in the BVDV-2 immunised animals were more varied (titres of <10±1280). Bolin and Ridpath (1995) have reported on the protective effect of passively acquired antibodies in calves challenged with a virulent isolate of BVDV. In their study, a titre of 16 prevented cell-free viraemia in serum, whereas, even at titres of 256, there was some virus dissemination in leukocytes. Our heterologous challenge experiments revealed a one-way cross-protection. All ewes immunised with BVDV-1 protected their foetuses from infection when challenged with BVDV-2, but none of the ewes immunised with BVDV-2 protected their foetuses from BVDV-1. These results could not be related to pre-challenge levels of antibody, since both groups of ewes had very similar, low levels of antibody to the respective challenge virus. For example, the ewe with the highest level of antibody to either challenge virus was F124 where a titre of 60 to BVDV-1 did not prevent foetal infection. A less marked degree of one-way cross-protection was observed by Vantsis et al. (1980) when conducting a similar experiment with BVDV-1 and BDV. In their study, heterologous protection against the BVDV-1 challenge was 50% compared to almost no protection (8%) following a heterologous BDV challenge. In the study of Bruschke et al. (1996) cross-protection was observed between two different BVDV-1 isolates, but was only measured in one direction. Considering protection and humoral antibody titres as indicators of immunogenicity, the BVDV-1 isolate used in this study seems to possess superior properties compared to the BVDV-2 isolate. The situation might be different for other strains, such as highly pathogenic BVDV-2 isolates, which may replicate more strongly in vivo. A one way cross-protection certainly suggests that the two viruses studied not only differ in antigenic classification, but also in their in vivo replicative patterns, and/or in the nature of the immune responses that they elicit. This implies that taxonomic assignments of BVD virus
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
195
isolates using serology or other techniques are not fully predictive of immunological cross-protection. This makes it difficult to draw firm conclusions regarding the significance of the antigenic variability and illustrates the difficulty of extrapolating from studies on specific viruses to generalisations on immunity in the field. Under natural conditions, it is likely that animals with a wide range of immunity will be exposed to different doses of an antigenically diverse array of pestiviruses, with differing potential for inducing foetal infection. Nevertheless, the results provide further evidence that exposure to one pestivirus does not ensure subsequent foetal cross-protection to another, even when the period between immunisation and challenge is less than 6 months. Furthermore, the pre-challenge level of neutralising antibody is not a reliable indication of the outcome, even where it is possible to test for homologous neutralisation in a test that utilises the challenge virus.
Acknowledgements Thanks to Christianne Bruschke for supplying the virus Pest-1/Bo/ID-DLO-4800/NL1993, and for helpful discussions. We are also grateful to our animal production unit for providing and maintaining the sheep, to Kathy Christiansen for statistical calculations and advice, and to Graham Heath for technical assistance. The work was supported by grant OD0326 from the Ministry of Agriculture, Fisheries and Food.
References Baker, J.C., 1995. The clinical manifestations of bovine viral diarrhoea infection. In: Baker, J.C., Houe, H. (Eds.), The Veterinary Clinics of North America: Food Animal Practice, vol. 11. Saunders, London, pp. 425±445. Bolin, S.R., Ridpath, J.F., 1995. Assessment of protection from systemic infection or disease afforded by low to intermediate titers of passively acquired neutralizing antibody against bovine viral diarrhoea virus in calves. Am. J. Vet. Res. 56, 755±759. Bruschke, C.J.M., van Rijn, P.A., Moormann, R.J.M., van Oirschott, J.T., 1996. Antigenically different pestivirus strains induce congenital infection in sheep: a model for bovine virus diarrhoea virus vaccine efficacy studies. Vet. Microbiol. 50, 33±43. Brownlie, J., Clarke, M.C., Hooper, L.B., Bell, G.D., 1995. Protection of the bovine fetus from bovine viral diarrhoea virus by means of a new inactivated vaccine. Vet. Rec. 137, 58±62. Carlsson, U., Alenius, S., Sundquist, B., 1991. Protective effect of an ISCOM bovine virus diarrhoea virus (BVDV) vaccine against an experimental BVDV infection in vaccinated and non-vaccinated pregnant ewes. Vaccine 9, 557±577. Corapi, W.V., French, T.W., Dubovi, E.J., 1989. Severe thrombocytopenia in young calves experimentally infected with noncytopathic bovine viral diarrhoea virus. J. Virol. 63, 3934±3943. Dekker, A., Wensvoort, G., Terpstra, C., 1995. Six antigenic groups within the genus pestivirus as identified by cross-neutralization assays. Vet. Microbiol. 47, 317±329. Duffell, S.J., Sharp, M.W., Winkler, C.E., Terlecki, S., Richardson, C., Done, J.T., Roeder, P.L., Herbert, C.N., 1984. Bovine virus diarrhoea mucosal disease virus-induced fetopathy in cattle: efficacy of prophylactic maternal pre-exposure. Vet. Rec. 114, 558±561. Harkness, J.W., Sands, J.J., Richards, M.S., 1978. Serological studies of mucosal disease virus in England and Wales. Res. Vet. Sci. 24, 98±103.
196
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
Harkness, J.W., Roeder, P.L., Drew, T., Wood, L., Jeffrey, M., 1987. In: Harkness, J.W. (Ed.), Pestivirus Infection of Ruminants. Brussels, CEC, 233 pp. Holm Jensen, M., 1981. Detection of antibodies against hog cholera virus and bovine viral diarrhoea virus in porcine serum. A comparative examination using CF PLA and NPLA assays. Acta Vet. Scand. 22, 85±94. Hyera, J.M.K., Liess, B., Frey, H.-R., 1987. A direct neutralising peroxidase-linked assay for detection and titration of antibodies to bovine viral diarrhoea virus. J. Vet. Med. B 34, 227±239. Jewett, C.J., Kelling, C.L., Frey, M.L., Doster, A.R., 1990. Comparative pathogenicity of selected bovine viral diarrhoea virus isolates in gnotobiotic lambs. Am. J. Vet. Res. 51, 1640±1644. Lambot, M., Douart, A., Joris, E., Letesson, J.-J., Pastoret, P.-P., 1997. Characterization of the immune response of cattle against non-cytopathic and cytopathic biotypes of bovine viral diarrhoea virus. J. Gen. Virol. 78, 1041±1047. Lobmann, M.S., Charlier, P., Klaasen, C.L., Zygraich, N., 1986. Safety of a temperature sensitive vaccine strain of bovine viral diarrhoea virus in pregnant cows. Am. J. Vet. Res. 47, 557±560. McGowan, M.R., Kirkland, P.D., Richards, S.G., Littlejohns, I.R., 1993. Increased losses in cattle infected with bovine pestivirus around the time of insemination. Vet. Rec. 133, 39±43. Meyling, A., Ronsholt, L., Dalsgaard, K., Jensen, A.M., 1987. In: Harkness, J.W. (Ed.), Pestivirus Infection of Ruminants. Brussels, CEC, 225 pp. Nettleton, P.F., 1987. Pathogenesis and epidemiology of border disease. Ann. Rech. Vet. 18, 147±155. Paton, D.J., Sands, J.J., Edwards, S., 1994. Border disease virus: delineation by monoclonal antibodies. Arch. Virol. 135, 241±252. Paton, D.J., 1995. Pestivirus diversity. J. Comp. Pathol. 112, 215±236. Paton, D.J., Sands, J.J., Lowings, J.P., Smith, J.E., Ibata, G., Edwards, S., 1995. A proposed division of the pestivirus genus using monoclonal antibodies, supported by cross-neutralisation assays and genetic sequencing. Vet. Res. 26, 92±109. Pellerin, C., van den Hurk, J., Lecomte, J., Tijssen, P., 1994. Identification of a new group of bovine viral diarrhoea virus strains associated with severe outbreaks and high mortalities. Virology 203, 260±268. Richardson, C., Herbert, C.N., Done, J.T., 1976. Experimental border disease in sheep: dose-response effect. Br. Vet. J. 132, 202±207. Ridpath, J.F., Bolin, S.R., Dubovi, E.J., 1994. Segregation of bovine viral diarrhoea virus into genotypes. Virology 205, 66±74. van Rijn, P.A., van Gennip, H.G.P., Leendertse, C.H., Bruschke, C.J.M., Paton, D.J., Moormann, R.J.M., van Oirschot, J.T., 1997. Subdivision of the pestivirus genus based on envelope glycoprotein E2. Virology 237, 337±348. Vantsis, J.T., Barlow, R.M., Gardiner, A.C., Linklater, K.A., 1980. The effects of challenge with homologous and heterologous strains of border disease virus on ewes with previous experience of the disease. J. Comp. Pathol. 90, 39±45. Vilcek, S., Nettleton, P.F., Paton, D.J., Belak, S., 1997. Molecular characterization of ovine pestiviruses. J. Gen. Virol. 78, 725±735.
Foetal cross-protection experiments between type 1 and type 2 bovine viral diarrhoea virus in pregnant ewes D.J. Paton*, G. Sharp, G. Ibata Central Veterinary Laboratory, Veterinary Laboratories Agency, Addlestone, Surrey KT15 3NB, UK
Abstract A flock of 82 non-pregnant ewes was split into three immunisation groups and given an intranasal dose of either cell culture medium, or a type 1 or a type 2 bovine viral diarrhoea virus (BVDV-1 or BVDV-2). Two months later the flock was reconstituted and after a further three weeks, the ewes were bred to pestivirus negative rams after synchronisation of oestrus using progesterone sponges. Fifty-five ewes were segregated into three challenge groups, each of which comprised ewes from different immunisation groups. At 7 weeks gestation, one challenge group was given an intranasal dose of cell culture medium, whilst the other two were given intranasal doses of either BVDV-1 or BVDV-2, using the same inocula as for the immunisations. Three weeks later, the ewes were killed and their foetuses tested for the presence of BVDV-1 and BVDV-2. The results showed that immunisation of six ewes without subsequent challenge did not lead to infection of any of their 11 foetuses. Challenge with BVDV-1 or BVDV-2 in the absence of immunisation lead to 15 out of 15 or 11 out of 14 foetuses becoming infected, respectively. Immunisation with the homologous virus to that used for challenge resulted in complete protection of 32 foetuses from 15 ewes. Heterologous protection was one way. All 12 foetuses from ewes immunised with BVDV-1 were protected from challenge with BVDV-2, whereas 18 foetuses from ewes immunised with BVDV-2 were all infected after challenge with BVDV-1. This provides evidence that a recent exposure to infection with one pestivirus does not necessarily induce foetal protection against another. The one-way result suggests that factors other than antigenic differences are involved in cross-protection. Crown Copyright # 1999 Published by Elsevier Science B.V. Keywords: Pestivirus; BVD; Sheep model; Foetal infection; Cross-protection
* Corresponding author. Tel.: +44-1932-357285; fax: +44-1932-357239; e-mail: [email protected] 0378-1135/99/$ ± see front matter Crown Copyright # 1999 Published by Elsevier Science B.V. PII S 0 3 7 8 - 1 1 3 5 ( 9 8 ) 0 0 2 6 9 - 7
186
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
1. Introduction Bovine viral diarrhoea virus (BVDV) is a small, enveloped, single stranded, positive sense RNA virus of the Pestivirus genus within the family Flaviviridae. Two types of BVDV (BVDV-1 and BVDV-2, synonym Pestivirus 1 and Pestivirus 2) can be distinguished from one another and from the two other principal pestivirus types, namely ovine border disease virus (BDV or Pestivirus 3) and classical swine fever virus (CSFV or Pestivirus 4). All of the pestiviruses are serologically related and many can infect a variety of artiodactyl hosts. Infection of cattle with BVDV is common in most parts of the world. Acute infection is characterised by a period of systemic viral infection and virus shedding, which is followed by recovery and development of long lasting antibody. Although acute infection is frequently unnoticed, it can cause or contribute to a variety of problems, including diseases of the reproductive, respiratory, and enteric systems (Baker, 1995). In cases of so-called virulent BVDV, there may be haemorrhagic disease and significant mortality in both young and adult animals. BVDV readily crosses the placenta and can cause infertility and embryonic or foetal death. Foetuses infected before the onset of immunocompetence can become persistently infected. Persistently infected (PI) animals may have reduced viability both in utero and after birth, but a proportion of PI animals survive, acting as an important reservoir of infection for other cattle. Prevention of foetal infection is, therefore, an important aim in BVD control and relies on either avoiding exposure of pregnant cattle to viral challenge, or immunisation before pregnancy. Many cattle develop an immunity to BVDV, either as a consequence of an earlier acute infection, or because of vaccination with available killed or live vaccines. It is known that prior infection or vaccination of the dam can protect foetuses from challenge in pregnancy with homologous (Duffell et al., 1984) or even somewhat dissimilar strains of BVDV (Brownlie et al., 1995), but there is also evidence that this protection may be incomplete (Harkness et al., 1987; Meyling et al., 1987). The antigenic and genetic diversity of BVDV is increasingly recognised. BVDV-1 has been the predominant type in most of the world, whereas BVDV-2 appears to be a newly emerging variant, so far only common in North America. Both types of BVDV can be associated with acute and persistent infections that range from being subclinical to fatal. However, in North America, the emergence of BVDV-2 has been associated with outbreaks of severe and sometimes haemorrhagic disease characterised by fatal acute infections (Pellerin et al., 1994; Ridpath et al., 1994). In Britain, the predominant pestivirus isolated from cattle has been BVDV-1 (Paton et al., 1995). A large proportion of adult cattle are seropositive due to earlier exposure to the virus (Harkness et al., 1978), and a killed vaccine based on BVDV-1 is starting to be used (Brownlie et al., 1995). However, serological cross-neutralisation between BVDV-1 and BVDV-2 is incomplete, both in vitro (Pellerin et al., 1994; Paton et al., 1995; Dekker et al., 1995) and in vivo (Bolin and Ridpath, 1995). There is, therefore, considerable interest to know whether the widespread immunity to BVDV-1 will protect pregnant cattle against challenge with BVDV-2. In the British sheep industry few measures are taken to control pestivirus infections. Characterisation of isolates has shown that two species of pestivirus predominate, namely
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
187
BVDV-1 and BDV, but that BVDV-2 has also been isolated (Paton et al., 1994; Vilcek et al., 1997). Previous studies have used sheep as a model to test the efficacy of BVDV immunisation and cross-protection (Vantsis et al., 1980; Carlsson et al., 1991; Bruschke et al., 1996). The aim of this study was to use sheep as a model to examine foetal crossprotection between infection with BVDV-1 and BVDV-2. 2. Materials and methods 2.1. Viruses Two non-cytopathic BVD viruses originally isolated from PI cattle were used for immunisation and challenge of sheep, and for in vitro serology. They were selected on the basis of their known antigenic and genetic classification (Paton, 1995; van Rijn et al., 1997), and for their proven ability to infect ovine foetuses (Bruschke et al., 1996). Virus Pest-1/Bo/ID-DLO-4800/NL-1993, a BVDV-1b isolated in The Netherlands in 1993 was kindly donated by Christianne Bruschke as a cell culture supernatant at passage one. It was passaged once more in bovine turbinate (BT) cells to give a stock with a titre of 105.6TCID50/ml. Pest-2/Bo/CVL-178003/GB-1987 is a BVDV-2 isolated in England in 1987. The stock used had been passaged three times in BT cells and had a titre of 107.8 TCID50/ml. The same inocula were used for both immunisation and challenge, each ewe receiving 2 ml of virus diluted to 105 TCID50/ml in MEM 10% FCS. 2.2. Experimental design A flock of 82 cross-bred ewes were used for the study. All were blood tested for pestivirus status using serology and attempted virus isolation prior to use. They were maintained on 20 ha of pasture throughout the experiment. Groups were segregated by placing them in securely fenced fields of approximately 4 ha, with an empty field between each group. Separate protective clothing was used for each group and animal handling and blood tests for each group were carried out on different days. The segregation of the flock into groups for immunisation and challenge is summarised in Fig. 1. In September, the flock was split randomly into three immunisation groups NI (no immunisation), IV1 (immunisation with BVDV-1) and IV2 (immunisation with BVDV-2). Each group comprised 26 principal ewes, but two additional ewes were left unimmunised in each of groups IV1 and IV2 in order to examine the efficiency of horizontal virus spread from the immunised animals. Immunisation consisted of an intranasal dose of either cell culture medium (MEM plus 10% FCS), BVDV-1 or BVDV2. A luer-lock syringe fitted with a spray nozzle was used to apply 1 ml of inoculum into each nostril. Blood samples were taken for serology approximately 3 and 6 weeks after immunisation. Six weeks after immunisation, the two unimmunised ewes within groups IV1 and IV2 were removed and housed in isolation elsewhere. At 8 weeks after immunisation the flock was reconstituted as a single group, and after a further 3 weeks, the ewes were bred to four rams after synchronisation of oestrus using
188
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
Fig. 1. Summary of cross-protection study design to show segregation of ewes into immunisation and challenge groups.
progesterone sponges. For this purpose, the ewes were split into five groups, each group being served by the rams during a single 24 h period on 5 consecutive days. One week later the ewes were divided into two groups, each of which was left with two rams for 3 weeks. Blood samples for serology and virology were taken from the rams immediately before their first introduction. At 5 weeks after breeding, ultrasound testing was used to select 55 ewes which were all at approximately the same stage of gestation. These ewes were segregated into three challenge groups: NC (no challenge), CV1 (challenge with BVDV-1) and CV2 (challenge with BVDV-2). The NC group comprised three ewes from each immunisation group. The CV1 and CV2 groups each comprised eight ewes from NI, eight ewes from IV1 and seven ewes from IV2 (Fig. 1). Ewes were assigned to challenge groups so as to match the range of antibody titres present in animals within each challenge group. At 7 weeks gestation, and 120 days after the initial immunisations, one challenge group was given an intranasal dose of cell culture medium, whilst the other two were given intranasal doses of either BVDV-1 or BVDV-2, using the same inocula and procedure as for the immunisations. Three weeks later, the ewes were killed and their foetuses tested for the presence of BVDV-1 and BVDV-2. For this purpose, an aseptic technique was used to collect umbilical blood, and a portion of spleen and kidney from each foetus. Foetal crown-rump lengths were recorded as an indicator of foetal age.
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
189
2.3. Test procedures A microtitre plate format of the neutralising peroxidase linked assay (NPLA) was used for serology (Hyera et al., 1987). Doubling dilutions of sera, starting at 1 in 10 were incubated for 1 h at 378C with equal volumes of medium containing approximately 100 TCID50 of either Pest-1/Bo/ID-DLO-4800/NL-1993 or Pest-2/Bo/CVL-178003/GB1987. BT cells were added and after 5 days further incubation, the presence of virus was tested for by peroxidase linked assay (PLA Holm Jensen, 1981) using a hyperimmune, bovine BVDV antiserum and a peroxidase linked rabbit anti-bovine conjugate (Dako). Titres are expressed as reciprocal dilutions of sera that neutralise 50% of replicate virus-infected monolayers. For calculation of geometric means, antibody titres were converted to logarithmic values and for this purpose, titres of <10 were given an arbitrary log value of 0.5. Virus isolation was attempted from pre-inoculation ewe sera, pre-breeding ram sera, and from post mortem collections of foetal and maternal sera, foetal spleen and kidney. Cell cultures and gamma irradiated foetal calf serum used for their growth had been previously screened and shown to be free of adventitious BVDV and BVDV inhibitors. Fresh serum samples in 10 ml aliquots were each applied directly to four wells of a microtitre plate, and then incubated with BT cells for 5 days at 378C. The presence of virus was detected by PLA, as described above. For each foetus, a pool of freshly collected spleen and kidney was homogenised and inoculated into flasks containing monolayers of rapidly growing BT cells. After 5 days incubation, the cultures were subpassaged into microtitre plate wells, incubated another 5 days and then fixed and immunostained for virus growth and detection as described for sera. In the case of isolation from tissues, additional immunostaining was carried out with monoclonal antibodies able to distinguish between the BVDV-1 and BVDV-2 viruses in the PLA. The monoclonal antibodies WB160 and WB166 are selective for Pest-1/Bo/ID-DLO-4800/ NL-1993, whilst WB437 and WB537 are selective for Pest-2/Bo/CVL-178003/GB-1987 (Paton et al., 1995). 2.4. Statistics The significance of differences in pregnancy rates and foetal infection rates between groups was analysed statistically by Fisher exact test. 3. Results 3.1. Response to immunisation All ewes were pestivirus negative and seronegative to both BVD viruses (titre of <10) when tested at 2 months prior to immunisation. All were seronegative at the time of immunisation. Ewes in the control NI group were still seronegative at the time of challenge, and those that received neither immunisation nor challenge remained seronegative throughout the experiment. All of the immunised ewes seroconverted
190
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
Table 1 Results of antibody tests and foetal virus tests on unchallengeda ewes from different immunisation groups Ewe
Immunisation
Antibody at challengea
Antibody at necropsy
BVDV-1
BVDV-1
BVDV-2
Foetuses infected
BVDV-2
H136 F127 F83
None (NI) None (NI) None (NI)
<10 <10 <10
<10 <10 <10
<10 <10 <10
<10 <10 <10
0/2 0/3 0/2
Mean
None (NI)
3
3
3
3
0%
G61 G289 F223
BVDV-1 (IV1) BVDV-1 (IV1) BVDV-1 (IV1)
1280 320 960
<10 <10 <10
1280 2560 1280
20 <10 <10
0/2 0/2 0/2
Mean
BVDV-1 (IV1)
741
3
1620
6
0%
E698 F256 G675
BVDV-2 (IV2) BVDV-2 (IV2) BVDV-2 (IV2)
10 <10 <10
80 320 480
<10 <10 <10
60 640 640
0/1 0/2 0/2
Mean
BVDV-2 (IV2)
5
229
3
293
0%
a
Antibody at challenge at time of challenge of other groups i.e. 120 days after immunisation.
within 6 weeks, and levels of neutralising antibody had reached a plateau by the time of challenge. The two uninoculated ewes within each of groups IV1 and IV2 had not seroconverted by the time that they were removed from the other ewes 6 weeks after the immunisation procedure. They were still seronegative 4 months later. Individual clinical examinations were not performed, but no gross signs of illness were observed in the ewes after immunisation. A single ewe from the IV2 group died of pneumonia 11 weeks after immunisation. Antibody levels at the time of challenge are shown in Tables 1±3. In general, higher titres were observed in ewes inoculated with BVDV-1 than in ewes inoculated with BVDV-2. The homologous, geometric mean antibody titre for all animals immunised with BVDV-1 was 1043. For ewes given BVDV-2, the equivalent homologous mean titre was 172. Antibody titres were much higher to the homologous virus than to the heterologous virus. Mean heterologous titres in the BVDV-1 and BVDV-2 immunised animals at the time of challenge were 6 and 5, respectively. 3.2. Ram testing and pregnancy diagnosis The four rams were pestivirus negative and seronegative when introduced to the first ewes. Ultrasound testing at approximately 5 weeks after breeding confirmed that 58 ewes were at the same stage of gestation. The proportions of ewes at this stage of pregnancy were not statistically different between immunisation groups: 20 out of 26, 17 out of 26 and 21 out of 25 for groups NI, IV1 and IV2, respectively.
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
191
Table 2 Results of antibody tests and foetal virus tests on ewes from different immunisation groups and challenged with BVDV-1 Ewe
Immunisation
(NI) (NI) (NI) (NI) (NI) (NI) (NI) (NI)
Antibody at challenge
Antibody at necropsy
BVDV-1
BVDV-1
<10 <10 <10 <10 <10 <10 <10 <10
BVDV-2
C161 D63 E619 F102 F492 G220 G510 G672
None None None None None None None None
<10 <10 <10 <10 <10 <10 <10 <10
Mean
None (NI)
3
3
D636 E253 E693 E757 F85 F221 G658
BVDV-1 BVDV-1 BVDV-1 BVDV-1 BVDV-1 BVDV-1 BVDV-1
(IV1) (IV1) (IV1) (IV1) (IV1) (IV1) (IV1)
2560 960 640 960 480 2560 2560
<10 <10 <10 20 <10 <10 <10
Mean
BVDV-1 (IV1)
1259
4
D631 E262 F124 F67 F91 G336 G357 G50
BVDV-2 BVDV-2 BVDV-2 BVDV-2 BVDV-2 BVDV-2 BVDV-2 BVDV-2
(IV2) (IV2) (IV2) (IV2) (IV2) (IV2) (IV2) (IV2)
<10 <10 60 <10 15 <10 <10 <10
Mean
BVDV-2 (IV2)
6
640 240 2560 320 120 60 240 960
Foetuses infected
BVDV-2 <10 <10 10 <10 <10 <10 <10 <10
2/2 1/1 2/2 2/2 3/3 2/2 2/2 1/1
355
4
100%
1280 640 960 1280 480 3840 2560
10 15 15 30 30 <10 15
0/2 0/3 0/2 0/2 0/2 0/2 0/4
1259
13
0%
120 160 480 160 480 120 40 240
3840 3840 15360 2560 10240 7680 3840 2560
5120 3840 1280 2560 20480 15360 2560 2560
3/3 2/2 2/2 2/2 2/2 3/3 2/2 2/2
178
5158
4340
100%
3.3. Effect of challenge No illness was observed in ewes after challenge. At post-mortem, all 55 ewes selected for the challenge study were found to be pregnant with one to three foetuses ranging in size from 12.5 to 18 cm of crown rump length. Gross foetal abnormalities were not apparent. Sera collected from ewes post-mortem (3 weeks after challenge) were all virus negative apart from one sample. This ewe (F125 ± Table 3) had been challenged with BVDV-2 without prior immunisation. A summary of the results of attempted virus isolation from foetal serum and organs is shown in Tables 1±3. Monoclonal antibody typing demonstrated that recovered virus was of the type used for challenge rather than immunisation. Eighteen foetuses obtained from nine ewes that had not been challenged in pregnancy were all virus negative, regardless of immunisation status. All 15 foetuses from eight unimmunised ewes that had been challenged with BVDV-1 were found to be infected, virus being isolated from both serum
192
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
Table 3 Results of antibody tests and foetal virus tests on ewes from different immunisation groups and challenged with BVDV-2 Ewe
Immunisation
(NI) (NI) (NI) (NI) (NI) (NI) (NI) (NI)
Antibody at challenge
Antibody at necropsy
BVDV-1
BVDV-1
<10 <10 <10 <10 <10 <10 <10 <10
BVDV-2
H100 D629 E421 F75 F125 G52 G487 G663
None None None None None None None None
<10 <10 <10 <10 <10 <10 <10 <10
Mean
None (NI)
3
3
E587 H440 F510 J321 F507 G333 G331
BVDV-1 BVDV-1 BVDV-1 BVDV-1 BVDV-1 BVDV-1 BVDV-1
(IV1) (IV1) (IV1) (IV1) (IV1) (IV1) (IV1)
960 7680 960 480 640 480 480
<10 20 <10 15 40 <10 20
Mean
BVDV-1 (IV1)
936
10
F111 E264 E686 F77 H776 F72 H541 H219
BVDV-2 BVDV-2 BVDV-2 BVDV-2 BVDV-2 BVDV-2 BVDV-2 BVDV-2
(IV2) (IV2) (IV2) (IV2) (IV2) (IV2) (IV2) (IV2)
20 <10 <10 <10 <10 <10 <10 <10
Mean
BVDV-2 (IV2)
4
<10 <10 <10 <10 <10 10 <10 15
Foetuses infected
BVDV-2 <10 <10 <10 40 <10 <10 <10 <10
0/1 0/1 1/1 2/2 2/2 2/2 2/2 2/3
4
4
79%
640 2560 640 480 640 320 320
<10 40 15 <10 20 <10 10
0/1 0/2 0/2 0/2 0/1 0/2 0/2
611
9
0%
640 80 120 960 1280 120 <10 60
10 <10 <10 <10 <10 <10 <10 <10
640 240 160 960 640 160 10 80
0/2 0/2 0/2 0/2 0/2 0/2 0/1 0/2
145
4
194
0%
and tissues in every case. A lesser proportion (11 out of 14) of foetuses were infected in the unimmunised ewes challenged with BVDV-2, and in some cases virus was recovered only from serum or only from tissues. This difference is not statistically significant, either on a ewe basis (p 0.47) or on a foetus basis (p 0.1). None of the foetuses were infected from ewes that had been challenged with the same virus as was used for prior immunisation. Likewise, none of the foetuses were infected from ewes previously immunised with BVDV-1 and then challenged with BVDV-2. However, all of the foetuses were infected from ewes that were immunised with BVDV-2 and then challenged with BVDV-1, and this difference is statistically significant. Serological results for samples collected at post mortem are shown in Tables 1±3. Unimmunised ewes challenged with BVDV-1 had seroconverted strongly when killed 3 weeks later (titre range 60±2560, Table 2), whereas unimmunised ewes that received BVDV-2 were mostly still seronegative at death (titre range <10±40, Table 3). Of the four pre-immunised groups that were challenged, either with a homologous or a heterologous
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
193
virus, a significant titre increase was only seen between challenge and death in one group. This was the group immunised with BVDV-2 and challenged with BVDV-1 (Table 2). At the final collection, these ewes had very high antibody titres to both BVDV1 and BVDV-2. 4. Discussion Use of sheep rather than cattle is much less costly for the in vivo study of crossprotection between pestiviruses, especially for work involving pregnant animals. Furthermore the predominant pestivirus types found in cattle also occur naturally in sheep, whereas the reverse is not true, since BDV is very rarely if ever isolated from cattle. A previous study had shown that the two bovine isolates Pest-1/Bo/ID-DLO-4800/ NL-1993 and Pest-2/Bo/CVL-178003/GB-1987 were able to infect ovine foetuses following intranasal exposure of their dams in early pregnancy (Bruschke et al., 1996). In the present experiment, a similar approach was used, except that sheep were maintained in groups at pasture rather than indoors in isolation units. This reduces costs further and may be less stressful for the animals. The results indicated that seronegative sheep kept in separate pastures on the same farm did not become infected from other inoculated groups. Indeed, horizontal transmission of virus did not occur from inoculated sheep to four control ewes kept in the same fields. Other reports have documented inefficient spread of ruminant pestiviruses between acutely infected ruminants and pastured sheep (Nettleton, 1987; McGowan et al., 1993). Intranasal inoculation of BVDV is a practical means to mimic the natural route of infection, whilst allowing for a controlled exposure to virus (McGowan et al., 1993). Selection of an appropriate dose properly requires prior titration of the inoculum in the target species (Richardson et al., 1976) and establishment of the normal level of field exposure. The dose chosen in the present study was the same as previously used by Bruschke et al. (1996) and reflects the level of virus shed from the nose of PI animals ± commonly in the region of 104±105 TCID50/ml. The BVDV-1 inoculum appeared to be more efficient than the BVDV-2 inoculum at infecting the foetuses of seronegative ewes (100% vs. 79% for the BVDV-2), although due to the relatively small group sizes the difference is not statistically significant. It also induced earlier seroconversion with higher titres of neutralising antibodies. Differences in immunogenicity may amongst other factors depend upon different patterns of viral replication, and this could also account for differences in foetal penetration. The majority of acute pestivirus infections in ruminants are mild or subclinical and little is known concerning differences in the replication or pathogenicity of the viruses involved. Isolates of extreme virulence or attenuation have been documented (Lobmann et al., 1986; Corapi et al., 1989), and subtler differences, for instance in the replication of bovine and ovine isolates in lambs, have also been reported (Jewett et al., 1990). Both of the viruses used in our inocula had been passaged only a small number of times in cell culture to minimise problems of attenuation. Both were also diluted so as to give the same titre of infectivity, as measured in bovine turbinate cell cultures, and titrations performed in ovine testicle cells gave similar results (data not shown). It would nevertheless be interesting to test higher doses
194
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
of the BVDV-2 inoculum in sheep, to see if this resulted in more extensive replication and a stronger seroconversion. Richardson et al. (1976) demonstrated a dose-dependent severity of congenital disease in lambs due to experimental pestivirus infection. With respect to immunogenicity, it has been shown that non-cytopathic biotypes of BVDV induce a stronger immunity than cytopathic viruses (Lambot et al., 1997), but otherwise there is not much information available on strain-specific differences. Jewett et al. (1990) described differences in the neutralising antibody responses of lambs to different pestivirus isolates, all given at the same dose. However, it was not clear how much these differences were due to pathogenic variability amongst the challenge viruses or to variable antigenic mismatch between the challenge viruses and the virus used for the neutralisation test. As expected, both groups of ewes that had been pre-immunised, did not transmit virus to their foetuses after a homologous virus challenge. Similar results have been reported for sheep, using BVDV and BDV (Vantsis et al., 1980) and for cattle using BVDV (Duffell et al., 1984), but this is the first such report concerning BVDV-2. However, in these experiments, the period between immunisation and challenge has been relatively short, and the duration of protective immunity has not been established. Our homologous challenge experiments give no indication of a particular titre of neutralising antibody that would equate to foetal protection. All of our BVDV-1 immunised sheep had moderately high titres of homologous neutralising antibody at the time of challenge (titres of 480), but the equivalent levels of antibody in the BVDV-2 immunised animals were more varied (titres of <10±1280). Bolin and Ridpath (1995) have reported on the protective effect of passively acquired antibodies in calves challenged with a virulent isolate of BVDV. In their study, a titre of 16 prevented cell-free viraemia in serum, whereas, even at titres of 256, there was some virus dissemination in leukocytes. Our heterologous challenge experiments revealed a one-way cross-protection. All ewes immunised with BVDV-1 protected their foetuses from infection when challenged with BVDV-2, but none of the ewes immunised with BVDV-2 protected their foetuses from BVDV-1. These results could not be related to pre-challenge levels of antibody, since both groups of ewes had very similar, low levels of antibody to the respective challenge virus. For example, the ewe with the highest level of antibody to either challenge virus was F124 where a titre of 60 to BVDV-1 did not prevent foetal infection. A less marked degree of one-way cross-protection was observed by Vantsis et al. (1980) when conducting a similar experiment with BVDV-1 and BDV. In their study, heterologous protection against the BVDV-1 challenge was 50% compared to almost no protection (8%) following a heterologous BDV challenge. In the study of Bruschke et al. (1996) cross-protection was observed between two different BVDV-1 isolates, but was only measured in one direction. Considering protection and humoral antibody titres as indicators of immunogenicity, the BVDV-1 isolate used in this study seems to possess superior properties compared to the BVDV-2 isolate. The situation might be different for other strains, such as highly pathogenic BVDV-2 isolates, which may replicate more strongly in vivo. A one way cross-protection certainly suggests that the two viruses studied not only differ in antigenic classification, but also in their in vivo replicative patterns, and/or in the nature of the immune responses that they elicit. This implies that taxonomic assignments of BVD virus
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
195
isolates using serology or other techniques are not fully predictive of immunological cross-protection. This makes it difficult to draw firm conclusions regarding the significance of the antigenic variability and illustrates the difficulty of extrapolating from studies on specific viruses to generalisations on immunity in the field. Under natural conditions, it is likely that animals with a wide range of immunity will be exposed to different doses of an antigenically diverse array of pestiviruses, with differing potential for inducing foetal infection. Nevertheless, the results provide further evidence that exposure to one pestivirus does not ensure subsequent foetal cross-protection to another, even when the period between immunisation and challenge is less than 6 months. Furthermore, the pre-challenge level of neutralising antibody is not a reliable indication of the outcome, even where it is possible to test for homologous neutralisation in a test that utilises the challenge virus.
Acknowledgements Thanks to Christianne Bruschke for supplying the virus Pest-1/Bo/ID-DLO-4800/NL1993, and for helpful discussions. We are also grateful to our animal production unit for providing and maintaining the sheep, to Kathy Christiansen for statistical calculations and advice, and to Graham Heath for technical assistance. The work was supported by grant OD0326 from the Ministry of Agriculture, Fisheries and Food.
References Baker, J.C., 1995. The clinical manifestations of bovine viral diarrhoea infection. In: Baker, J.C., Houe, H. (Eds.), The Veterinary Clinics of North America: Food Animal Practice, vol. 11. Saunders, London, pp. 425±445. Bolin, S.R., Ridpath, J.F., 1995. Assessment of protection from systemic infection or disease afforded by low to intermediate titers of passively acquired neutralizing antibody against bovine viral diarrhoea virus in calves. Am. J. Vet. Res. 56, 755±759. Bruschke, C.J.M., van Rijn, P.A., Moormann, R.J.M., van Oirschott, J.T., 1996. Antigenically different pestivirus strains induce congenital infection in sheep: a model for bovine virus diarrhoea virus vaccine efficacy studies. Vet. Microbiol. 50, 33±43. Brownlie, J., Clarke, M.C., Hooper, L.B., Bell, G.D., 1995. Protection of the bovine fetus from bovine viral diarrhoea virus by means of a new inactivated vaccine. Vet. Rec. 137, 58±62. Carlsson, U., Alenius, S., Sundquist, B., 1991. Protective effect of an ISCOM bovine virus diarrhoea virus (BVDV) vaccine against an experimental BVDV infection in vaccinated and non-vaccinated pregnant ewes. Vaccine 9, 557±577. Corapi, W.V., French, T.W., Dubovi, E.J., 1989. Severe thrombocytopenia in young calves experimentally infected with noncytopathic bovine viral diarrhoea virus. J. Virol. 63, 3934±3943. Dekker, A., Wensvoort, G., Terpstra, C., 1995. Six antigenic groups within the genus pestivirus as identified by cross-neutralization assays. Vet. Microbiol. 47, 317±329. Duffell, S.J., Sharp, M.W., Winkler, C.E., Terlecki, S., Richardson, C., Done, J.T., Roeder, P.L., Herbert, C.N., 1984. Bovine virus diarrhoea mucosal disease virus-induced fetopathy in cattle: efficacy of prophylactic maternal pre-exposure. Vet. Rec. 114, 558±561. Harkness, J.W., Sands, J.J., Richards, M.S., 1978. Serological studies of mucosal disease virus in England and Wales. Res. Vet. Sci. 24, 98±103.
196
D.J. Paton et al. / Veterinary Microbiology 64 (1999) 185±196
Harkness, J.W., Roeder, P.L., Drew, T., Wood, L., Jeffrey, M., 1987. In: Harkness, J.W. (Ed.), Pestivirus Infection of Ruminants. Brussels, CEC, 233 pp. Holm Jensen, M., 1981. Detection of antibodies against hog cholera virus and bovine viral diarrhoea virus in porcine serum. A comparative examination using CF PLA and NPLA assays. Acta Vet. Scand. 22, 85±94. Hyera, J.M.K., Liess, B., Frey, H.-R., 1987. A direct neutralising peroxidase-linked assay for detection and titration of antibodies to bovine viral diarrhoea virus. J. Vet. Med. B 34, 227±239. Jewett, C.J., Kelling, C.L., Frey, M.L., Doster, A.R., 1990. Comparative pathogenicity of selected bovine viral diarrhoea virus isolates in gnotobiotic lambs. Am. J. Vet. Res. 51, 1640±1644. Lambot, M., Douart, A., Joris, E., Letesson, J.-J., Pastoret, P.-P., 1997. Characterization of the immune response of cattle against non-cytopathic and cytopathic biotypes of bovine viral diarrhoea virus. J. Gen. Virol. 78, 1041±1047. Lobmann, M.S., Charlier, P., Klaasen, C.L., Zygraich, N., 1986. Safety of a temperature sensitive vaccine strain of bovine viral diarrhoea virus in pregnant cows. Am. J. Vet. Res. 47, 557±560. McGowan, M.R., Kirkland, P.D., Richards, S.G., Littlejohns, I.R., 1993. Increased losses in cattle infected with bovine pestivirus around the time of insemination. Vet. Rec. 133, 39±43. Meyling, A., Ronsholt, L., Dalsgaard, K., Jensen, A.M., 1987. In: Harkness, J.W. (Ed.), Pestivirus Infection of Ruminants. Brussels, CEC, 225 pp. Nettleton, P.F., 1987. Pathogenesis and epidemiology of border disease. Ann. Rech. Vet. 18, 147±155. Paton, D.J., Sands, J.J., Edwards, S., 1994. Border disease virus: delineation by monoclonal antibodies. Arch. Virol. 135, 241±252. Paton, D.J., 1995. Pestivirus diversity. J. Comp. Pathol. 112, 215±236. Paton, D.J., Sands, J.J., Lowings, J.P., Smith, J.E., Ibata, G., Edwards, S., 1995. A proposed division of the pestivirus genus using monoclonal antibodies, supported by cross-neutralisation assays and genetic sequencing. Vet. Res. 26, 92±109. Pellerin, C., van den Hurk, J., Lecomte, J., Tijssen, P., 1994. Identification of a new group of bovine viral diarrhoea virus strains associated with severe outbreaks and high mortalities. Virology 203, 260±268. Richardson, C., Herbert, C.N., Done, J.T., 1976. Experimental border disease in sheep: dose-response effect. Br. Vet. J. 132, 202±207. Ridpath, J.F., Bolin, S.R., Dubovi, E.J., 1994. Segregation of bovine viral diarrhoea virus into genotypes. Virology 205, 66±74. van Rijn, P.A., van Gennip, H.G.P., Leendertse, C.H., Bruschke, C.J.M., Paton, D.J., Moormann, R.J.M., van Oirschot, J.T., 1997. Subdivision of the pestivirus genus based on envelope glycoprotein E2. Virology 237, 337±348. Vantsis, J.T., Barlow, R.M., Gardiner, A.C., Linklater, K.A., 1980. The effects of challenge with homologous and heterologous strains of border disease virus on ewes with previous experience of the disease. J. Comp. Pathol. 90, 39±45. Vilcek, S., Nettleton, P.F., Paton, D.J., Belak, S., 1997. Molecular characterization of ovine pestiviruses. J. Gen. Virol. 78, 725±735.