PPe

MOa

b

a

ARRAA

KVBVL

1

sdgcssaVa

df

0d

Vaccine 30 (2012) 2131– 2139

Contents lists available at SciVerse ScienceDirect

Vaccine

jou rn al h om epa ge: www.elsev ier .com/ locate /vacc ine

rotective efficacy of Bluetongue virus-like and subvirus-like particles in sheep:resence of the serotype-specific VP2, independent of its geographic lineage, isssential for protection

. Stewarta, C.I. Dovasb, E. Chatzinasioub, T.N. Athmarama, M. Papanastassopouloub,. Papadopoulosb, P. Roya,∗

Department of Infectious Diseases, London School of Hygiene and Tropical Medicine, United KingdomLaboratory of Microbiology and Infectious Diseases, Faculty of Veterinary Medicine, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece

r t i c l e i n f o

rticle history:eceived 29 November 2011eceived in revised form 10 January 2012ccepted 13 January 2012vailable online 26 January 2012

eywords:accineluetongue VirusLPsineage

a b s t r a c t

There have been multiple separate outbreaks of Bluetongue (BT) disease of ruminants in Europe since1998, often entering via the Mediterranean countries of Italy, Spain and Greece. BT is caused by anorbivirus, Bluetongue virus (BTV), a member of the family Reoviridae. BTV is a non-enveloped double-capsid virus, which encodes 7 structural proteins (VP1-VP7) and several non-structural proteins (NS1,NS2, NS3/3a and NS4) from ten double-stranded RNA segments of the genome. In this report, we haveprepared BTV virus-like particles (VLPs, composed of VP2, VP3, VP5 and VP7) and sub-viral, inner core-like particles (CLPs, VP3 and VP7) using a recombinant baculovirus expression system. We compared theprotective efficacy of VLPs and CLPs in sheep and investigated the importance of geographical lineagesof BTV in the development of vaccines. The Greek crossbred Karagouniko sheep, which display mild tosub-clinical BT, were vaccinated with VLPs or CLPs of BTV-1, derived from western lineage and were chal-lenged with virulent BTV-1 from an eastern lineage. All VLP-vaccinated animals developed a neutralisingantibody response to BTV-1 from both lineages prior to challenge. Moreover, post-challenged animals

had no clinical manifestation or viraemia and the challenged virus replication was completely inhibited.In contrast, CLP-vaccinated animals did not induce any neutralising antibody response but developed thegroup specific VP7 antibodies. CLPs also failed to prevent the clinical manifestation and virus replication,but in comparison to controls, the severity of disease manifestation and viraemia was mitigated. The datademonstrated that the outer capsid was essential for complete protection, while the geographical originof the BTV was not critical for development of a serotype specific vaccine.

. Introduction

Bluetongue (BT) is one of the most important diseases of live-tock. The disease is caused by Bluetongue virus (BTV), a segmented,ouble-stranded (ds) RNA virus, which is a member of the orbivirusenus within the Reoviridae family. BTV is a non-enveloped virusonsisting of two capsids (outer and inner) that enclose three tran-criptase complex proteins (VP1, VP4 and VP6) and 10 dsRNAegments (S1-S10) of the genome [1]. The inner capsid (termed

s “core”) has two concentric layers: the first layer is formed byP3, one of the four major viral proteins, and the second by VP7,nother major protein. Both proteins are highly conserved across

∗ Corresponding author at: London School of Hygiene and Tropical Medicine, Lon-on WC1E 7HT, United Kingdom. Tel.: +44 2077 927 2324;ax: +44 207 637 4314.

E-mail address: polly.roy@lshtm.ac.uk (P. Roy).

264-410X/$ – see front matter © 2012 Elsevier Ltd. All rights reserved.oi:10.1016/j.vaccine.2012.01.042

© 2012 Elsevier Ltd. All rights reserved.

all BTV serotypes. The surface layer of the core, which is formed byVP7, acts as a scaffold for the deposition of other two major cap-sid proteins, VP2 and VP5 that form the outer capsid [1]. VP2 isthe most exposed protein of the virus; it is the most variable pro-tein of the seven proteins and is the determinant of the serotypespecificity of BTV. While VP2 is responsible for attachment of thevirus to the cell surface, VP5 is involved in membrane penetrationand together they are responsible for virus entry into the host cells.BTV is endemic in many parts of the world and to date, 26 dis-tinct serotypes (BTV-1, -2, -3, etc.) have been recognized aroundthe world, mainly based on the neutralisation activity of VP2. BTVis transmitted from animal to animal by biting midges (Culiciodesspp.), thus the spread of BTV coincides with the distribution of theinsect vectors.

In Europe, BTV outbreaks have been sporadic since the firstmajor incidence in Spain and Portugal in 1956, which lasted until1960. However, since 1998 BTV incursion became more frequent,initially restricted to the southern fringe of Europe until 2006

2 ccine 3

w9ahdtasssbcisrapalTB[acels[td

uasdv[oobmMosvVsdBtsaag

besceaiiocc

tanide ISA 206 VG consisting of mannide oleate and mineral oil)respectively. Groups I and II were vaccinated with either BTV CLPs

132 M. Stewart et al. / Va

hen the first outbreaks in northern Europe occurred. At least serotypes have entered Europe and resulted in high morbiditynd mortality in livestock. The economic losses associated with BTave been of considerable significance in recent years, not onlyue to animal infection but also due to restrictions imposed byhe International Animal Health Organization on animal trade andnimal movement from regions where the virus is endemic. Theeverity of the disease is often associated with certain breeds ofheep, in particular, the fine wool and mutton producing Europeanheep breeds. In comparison, infection in other sheep breeds (e.g.,reeds of Greek Islands, Lesbos and Friesen) and also in goats andattle, results in subclinical or mild response with high circulat-ng virus titres [2]. These animals are potential reservoirs for thepread of the virus. Indeed, Greece is considered to be a majoroute of BTV entry into Europe since 4 serotypes (BTV-1, -4, -9nd -16) were first detected in Greece prior to spread in otherarts of Europe. It was clear from the molecular epidemiologicalnalysis of European isolates that there were two distinct evo-utionary lineages, eastern and western, regardless of serotypes.his was highlighted in particular by the separate outbreaks ofTV-1 into the Mediterranean basin via Greece in 2001 and 20063] and via the Iberian Peninsula in 2007–2010. The sequencenalyses of virus genome segment 3, which encodes the highlyonserved VP3, from the Greek isolate (GRE2001/01) exhibitedastern lineage (or topotype) isolate of BTV, while the virus iso-ates from the Iberian Peninsula (ALG2006/01 and MOR2006/06)howed that the outbreaks were due to western lineage isolates4]. The importance of the evolutionary difference between thesewo lineages in vaccine development has not been addressed toate.

Vaccination is the most effective means to control BT partic-larly if the vaccine affords complete protection to the animalss well as inhibits replication, thereby breaking the transmis-ion cycle between insect vectors and animal hosts. A number ofifferent candidate vaccines have recently been developed witharying degrees of success, including attenuated, inactivated (killed5]), pox-based [6,7] and recombinant protein based [8–11]. Onef the successful vaccine approaches has been the productionf BTV virus-like particles (VLPs) synthesized by a recombinantaculovirus expression system [8,12,13]. VLPs are non-infectiousimics of the virus structure as they contain no genetic material.oreover, VLPs are made of only four (VP2, VP3, VP5 and VP7)

f the seven BTV structural proteins and none of the BTV non-tructural (NS) proteins. Thus, unlike live attenuated and killedirus vaccines that are currently used to control BTV infection,LP-vaccinated animals do not elicit antibodies for the remainingtructural or NS proteins. These criteria are directly relevant foristinguishing between infected and vaccinated animals (DIVA).TV VLPs consist of two capsids, while the viral core-like par-icles (CLPs) are formed only by VP3 and VP7. While extensivetudies have been carried out on protective efficacies of VP2, VP5nd VLPs in vaccinated animals, the protective efficacy of VP3nd VP7 (CLPs) lacking the VP2 and VP5, have not been investi-ated.

In this report, we have compared the protective efficacy affordedy BTV-1 (RSA strain) VLPs and CLPs derived from a western lin-age and challenged with eastern lineage virulent BTV-1 (Greecetrain). The results obtained demonstrate that an intact outerapsid consisting of VP2 and VP5 is crucial to prevent BT dis-ase and viraemia in sheep. In contrast, CLPs were not able tofford complete protection from disease but mitigated the sever-ty of the disease. In addition, viraemia was also slightly reducedn CLP vaccinated animals. Our results also indicate that therigin of the virus isolate, either eastern or western, is not a

ritical factor when developing a serotype specific candidate vac-ine.

0 (2012) 2131– 2139

2. Material and methods

2.1. Virus stocks, genetic material and cells

BSR cells (derived from baby hamster kidney cells) werecultured in Dulbecco’s modified Eagle medium (DMEM, Lonza)supplemented with 5% foetal calf serum (FCS, PAA). BTV-1 RSAr-rrr/01 laboratory strain (BTV-1 RSA; western lineage) and BTV-1GRE2001/07 isolate (eastern lineage) virus stocks were propagatedin BSR cells at a multiplicity of infection (MOI) of 0.1 and harvestedwhen 95–100% cytopathic effect was evident. Titres of viral stockswere obtained by plaque assay.

The BTV-1 GRE2001/07 isolate from naturally infected sheep,was inoculated into a donor animal. Blood was collected 8 dayspost-infection, washed twice in equal volume of PBS (pH 7.2),aliquoted and stored at −80 ◦C for use as a challenge inoculum. Thechallenge virus was isolated from Rodos Island, a geographic areato the east and southeast of Greece, which encompasses Cyprus andthe Middle East, and based on the phylogenetic analysis of Seg-3(acc. number: DQ186807) belongs to the eastern topotype [3,14].

Recombinant baculoviruses that expressed BTV proteins for BTVVLP and CLP production were grown in Spodoptera frugiperda (Sf9)insect cells. Sf9 were cultured at 28 ◦C in Insect-Xpress (Lonza)in the presence of penicillin (100 U/ml), streptomycin (100 �g/ml)and Amphotericin B (250 ng/ml; Lonza).

2.2. BTV-1 RSA VLP and CLP vaccines

Core-like particles (CLPs) were produced from the co-expressionof BTV VP3 and VP7 from a single recombinant baculovirus [15]. Sf9were infected with an MOI of 3–5, harvested 60 h post-infection andCLPs were purified from infected cells as described previously [15].BTV-1 RSA VLP were generated by the expression of BTV-1 VP2, VP3,VP5 and VP7 from a single baculovirus in Sf9 insect cells infectedat MOI of 3–5 as described by Noad et al. [16]. VLPs were purifiedfrom infected Sf9 cells as described [8,17,18]. Purified BTV-1 VLPswere aliquoted and trehalose was added to a final concentration of5% (w/v) to preserve the integrity of the particles and stored at 4 ◦Cprior to vaccination as described [8].

2.3. Electron microscopy

Formavar carbon coated grids were glow discharged, treatedwith 1% (w/v) Alcian blue, and rinsed with sterile deionised water.Purified VLPs or CLPs were applied to the grids and stained with2% (w/v) phosphotungstic acid (PTA) pH 6.8. The grids were air-dried for 20 min and examined using a Jeol 1200 EX transmissionmicroscope.

2.4. BTV-1 RSA VLP and BTV CLP vaccination and virus challengeof sheep

To assess the efficacy of the vaccine in the target species, 21female Karagouniko crossbred sheep (18 months old) were used.The animals were free of respiratory, digestive, umbilical, para-sitic and osteo-articular diseases; and were seronegative againstBTV and Q fever. The animals were acclimatised to the insect-freehousing and rectal temperatures were recorded for 2 weeks prior tovaccination. They were fed hay and concentrate and supplied waterad libidum. Three groups of 6–7 sheep were segregated at random.Group III was the control group, divided into 3 and 4 animals, whichwere inoculated with saline or saline plus SEPPIC adjuvant (Mon-

or BTV-1 RSA VLPs (∼50–100 �g) that had been mixed with SEPPICadjuvant as per the manufacturer’s recommendation.

cine 3

oawanB

2e

dsdBcaoraR

2

aaaasSV

Aci[acw

2

tvasdwtcldp(fst

2(

fi

M. Stewart et al. / Vac

On day 0 the animals were vaccinated by subcutaneous routen the lateral side of the thorax, and on day 26 animals received

second vaccination. Similarly, the control groups were boostedith the saline or saline plus adjuvant on day 26. On day 53, all the

nimals were challenged with a 1 ml subcutaneous injection to theeck and 1 ml intradermally in the inner side of thigh with GreekTV-1 isolate (GRE2001/07, 2 × 104 pfu/ml).

.5. Clinical observations of infected sheep and post mortemxamination

Sheep were observed after challenge with virulent BTV for theevelopment of classical bluetongue disease symptoms and allymptoms were recorded. The rectal temperature was measuredaily from days 52 until the end of the trial as an indicator ofTV infection. All animals were euthanized thirty days after thehallenge and subjected to full necropsies. Animals were treatedccording to the ethical rules of National and European regulationsn animal welfare. All animal experiments were approved by theelevant institutional animal ethics committee and carried out inccordance with approved guidelines (National Health and Medicalesearch Council, Greece, 2009).

.6. Serological response to the vaccines

Blood samples were collected by jugular puncture in plain tubesnd sera were collected to monitor the antibody response to BTVntigens on days 0, 21, 41, 53, 65, 75 and prior to euthanasia (day 83nd onwards). The collected sera were heat inactivated for 30 mint 56 ◦C. The sera were tested for an antibody response to the grouppecific antigen VP7 using a commercial competition ELISA kit (IDCREEN® Blue Tongue Competition Kit; ID VET). The response toP7 was determined as per the manufacturer’s protocols.

The neutralising antibody response in all animals was assessed. comparison of the neutralising titres 14 days after the secondhallenge using both BTV-1 RSArrrr/01 and GRE2001/07 isolatesn a serum neutralisation 50TCID50 were performed as described19,20]. Plaque reduction serum neutralisation were performed forll serum samples using the GRE2001/07 isolates over the timeourse of the experiment (days 0, 21, 41, 53, 65, 75, 83). All dilutionsere preformed in triplicate and the assay repeated at least twice.

.7. Quantification of BTV dsRNA by real-time RT-PCR

Whole blood was collected from sheep in ethylene diaminte-raacetic acid (EDTA) tubes to enable the detection of the circulatingirus. Samples were collected on days 0, 3, 5, 8, 12, 22 and 30,fter challenge. The viral RNA was extracted from collected bloodamples and quantified by reverse transcription real-time PCR asescribed [21]. Briefly, the viral RNA was extracted from 200 �lashed red blood cells using the in-house method B based on

he lysing and nuclease-inactivating properties of a guanidinehaotropic agent and a phenol–chloroform extraction step, fol-owed by selective binding of RNA on a silica membrane. The dsRNAenaturation, cDNA synthesis, and quantitative real-time PCR waserformed using in parallel an exogenous dsRNA internal controlIC) from Epizootic hemorrhagic disease virus (EHDV) to monitoror dsRNA denaturation and RT efficiency. BTV dsRNA copies of theamples tested were calculated according to a standard curve andhe measured Ct values and finally adjusted to the recovery of EHDV.

.8. Quantification of infectious BTV by embryonated chicken egg

ECE) inoculation

Serial 10-fold dilutions of washed RBC were inoculated intove 11–12-day-old embryonated chicken eggs [22–24]. All chicken

0 (2012) 2131– 2139 2133

embryos that died between 2 and 7 days post-inoculation showedsubcutaneous haemorrhages consistent with BTV infection. Forconfirmation, tissues from embryos of the highest positive dilu-tion were frozen, homogenised in 10 volumes of Eagles medium(Invitrogen) and inoculated onto BHK-21 cells for virus detection.Virus titre was calculated by the Reed–Muench method.

2.9. Sequence and phylogenetic analysis of VP2 fromGRE2001/07, challenge virus

BSR were infected with the challenge virus BTV-1 GRE2001/07at a multiplicity of infection (MOI) of 1 and viral dsRNA was purifiedfrom the infected cells as previously described [25]. An exact copyof segment 2 (S2) from viral dsRNA was generated in a sequence-independent manner using the method of full-length amplificationof cDNAs (FLAC [26]) and cloned for sequencing. The full-lengthsequence of the VP2 coding region was generated using a seriesof overlapping primers and assembled in BioEdit. The nucleotidesequence of BTV-1 GRE S2 has been deposited into GenBank (acces-sion number JN635334).

Additional published complete VP2 sequences of BTV-1, -2and -8 were used for phylogenetic analysis. MEGA 5.0 softwarewas used to infer phylogenies. The sequences were aligned andthe optimal amino acid substitution model was determined usingAkaike Inforation Criteria (AIC) and a Likelihood Ratio Test (LRT).Maximum-likelihood trees were constructed using the followingparameters: JTT + G model (Jones–Taylor–Thornton model; discreteGamma distribution with 5 rate categories) and Nearest-NeighbourInterchange heuristic tree interfere method. The reliability of thephylogenetic hypothesis was assessed using non-parametric boot-strap with 1000 replications.

2.10. Statistical analysis

Differences between the neutralising antibody titres of the vac-cinated and non-vaccinated groups of sheep after challenge wereanalysed using non-parametric Mann–Whitney test for indepen-dent groups [27]. Statistical significance of the tests was based onp values equal or lower than 0.05.

3. Results

3.1. Preparation of BTV vaccine candidates and preparation ofchallenge virus

For this study, both VLP and CLP vaccines were derivedfrom a BTV-1 RSA isolate (western lineage) as described previ-ously [15,16]. The proteins were expressed from different loci inthe baculovirus genome and the particles purified as previouslydescribed. The morphology of the particles were analysed by elec-tron microscopy and showed that the structures of the VLPs andCLPs were correct and intact (data not shown). Furthermore, pro-tein integrity was monitored by SDS-PAGE, which showed thatthere was no degradation of the outer capsid proteins in the VLPpreparation.

In order to determine if a vaccine originating from a western lin-eage could protect animals challenged with a virulent virus isolatefrom the same serotype but with a different geographical lineage,a challenge virus was prepared from a BTV-1 strain that was iso-lated in Greece (i.e., eastern lineage). Since the total virulence ofthe tissue culture amplified BTV can only be revived in sheep, thechallenge virus was passaged in sheep prior to use for challenge.

To confirm the lineage of the vaccine strain versus the challengestrain, the coding region of segment 2 (VP2) was sequenced for bothBTV-1 isolates. Phylogenetic analysis of VP2 from the BTV-1 RSAand GRE2001/07 confirmed that the viruses originated from two

2134 M. Stewart et al. / Vaccine 30 (2012) 2131– 2139

Fig. 1. Phylogenetic tree inferred with maximum likelihood analysis, using complete VP2 amino acid sequences of BTV and the JTT + G model. The numbers indi-c as roon late us

daataais[

3v

3BIpstbvctnv(bdr

ated on branches are non-parametric bootstrap (NPB) probabilities. The tree wumber–origin–serotype. The challenge GRE2001/07 virus and the South Africa iso

ifferent lineages, which correspond to the eastern (Australasian)nd western (North American/South African) groups (Fig. 1). Themino acid sequence similarity varied from 92.8% to 99.2% withinhe eastern lineage, from 93.2% to 99.9% within the western lineagend from 77.3% to 80.9% between the two lineages. The VP2 aminocid sequence similarity between the BTV-1 RSA and GRE2001/07solate was 82.1%. Although there were changes to the amino acidequence, the two regions with the neutralisation epitopes in VP228,29] were conserved in sequence of both BTV-1 isolates (Fig. 2).

.2. Immunological response to BTV-1 VLPs and CLPs inaccinated sheep

Twenty-one Karagouniko cross female sheep were divided into groups (I, II and III) and immunized with BTV CLPs (Group I),TV-1 RSA VLPs (Group II) or saline/saline plus adjuvant (Group

II). All animals were vaccinated twice on days 0 and 26 using arime and boost regime as described in Section 2. Each vaccinatedheep received ∼50–100 �g of either VLPs or CLPs. Development ofhe BTV antibody response in vaccinated animals was monitoredased on VP7, the group specific antigen using an ELISA. Animalsaccinated with BTV-1 VLPs recognized BTV VP7 after the first vac-ination from day 21 onwards. All vaccinated sheep had a responseo VP7 after the second vaccination (41 days; Fig. 3). As expected,one of the control animals had a response to BTV VP7 prior toirulent BTV-1 challenge (day 0, Fig. 3). In the vaccinated groups

Groups I and II), the antibody response to VP7 remained fairly sta-le and over the positive threshold until the end of the study. Atay 80, 29 days after challenge all sheep had elicited an antibodyesponse to VP7.

ted at midpoint as implemented in MEGA version 5. Isolate abbreviations: acc.ed for VLP construct are indicated in bold.

3.3. VLPs but not CLPs vaccinated animals developed neutralisingantibodies to BTV-1

Although ELISA to detect antibodies to VP7 in vaccinated ani-mals to the VLPs and CLPs was indicative of an immune response, itdid not demonstrate if a neutralising antibody response had devel-oped. The neutralising antibody titres against both BTV-1 RSA andBTV-1 GRE2001/07 were therefore undertaken to compare animalsera collected at day 14 after the second vaccination and prior to vir-ulent virus challenge. This would also determine if the amino acidchanges observed in the VP2 would influence the ability of the virusto be neutralised. All animals in Group II immunized with BTV-1VLPs produced neutralising antibodies against both BTV-1 isolates(Fig. 4). The antibody titres were variable and ranged from 32 to128 (the upper limit tested). Only animals that were vaccinatedwith BTV-1 VLPs elicited neutralising antibodies prior to challenge.In contrast to VLP vaccinated sheep sera, sera of the CLP vaccinatedanimals did not develop any neutralising antibody against BTV-1,similar to the sera of the control animals (Group III; Fig. 4). Theseresults confirmed that the two outer capsid proteins are involvedin the triggering of the neutralising antibody response.

For neutralising antibody titre, the conventional 50% plaquereduction assay was undertaken using the challenge virus isolateover the time course of the experiment (Table 1). None of the ani-mals had any neutralising antibodies prior to vaccination. All sheepin Group II (VLP vaccinated) after the first vaccination (day 26)developed neutralising antibody titres between 16 and 32. After

the second vaccination the antibody titres in this group of ani-mals increased steadily from 32 to 128 on day 41 (15 days afterboost) and from 64 to 128 on day 53 (challenge day, mean titre84.48 ± 1.41, p < 0.05; Table 1). However, none of the animals in

M. Stewart et al. / Vaccine 30 (2012) 2131– 2139 2135

F Africw n regi

Gvwa

FcVeVr

ig. 2. Prediction of continuous B cell epitopes within the VP2 of the BTV-1 Southhich differ in the BTV-1 GRE2001/07 challenge virus are shaded. The neutralisatio

roup I (CLP vaccinated), or Group III (control, saline plus adju-ant) had any neutralising antibody prior to challenge and all titresere less than 4, the threshold of the assay. This was expected,

s the animals had not been exposed to VP2, the protein that

ig. 3. Group specific immune response of sheep vaccinated with VLPs. Sero-onversion of vaccinated sheep was monitored using competitive ELISA specific toP7, the BTV group specific core antigen. The average % competition is presented forach group, CLP vaccinated (�), VLP vaccinated (�) and control animals (© and �).1 and V2 indicate the day sheep were vaccinated and boosted with immunogen,espectively. C indicates the day sheep were challenged with virulent BTV.

a isolate (acc. number: ACR58459) using COBEpro software. VP2 amino acid sites,ons R1 and R2 proposed by DeMaula et al. [29] are boxed.

contains the viral neutralisation epitopes. The antibody titres of thepost-challenge Group II animal sera continued to increase, with themaximum titre of 1024 recorded on day 83 (mean titre 1024 ± 0.00,p < 0.05). Also Groups I and III developed neutralising antibodies

Fig. 4. Neutralisation antibody response of sheep vaccinated with VLPs against BTV-1 from two different lineages. Virus neutralisation antibody titres against BTV-1RSA and BTV-1 Greece of individual sheep vaccinated with BTV-1 RSAVLP and BTV-CLP and non-vaccinated sheep prior to challenge with virulent virus as determinedby the neutralisation of 50 TCID50. The group each animal belongs to is indicatedunderneath the animal number neutralising antibody titres against BTV-1 RSA areindicated by the white bar and BTV-1 Greece is indicated by the black bars. Titresbelow the threshold of the assay (dashed line) were considered to be negative (below8).

2136 M. Stewart et al. / Vaccine 30 (2012) 2131– 2139

Table 1Neutralising antibody titres raised in sheep vaccinated with either CLPs (Group I) or VLPs (Group II) and control animals (Group III) determined by serum neutralisationplaque reduction assay using BTV-1 GRE2001/07. Serum samples were taken during the vaccination period and after challenge with virulent BTV-1 GRE2001/07 isolate.Neutralising antibody titres for each animals and the mean for each group is presented. The neutralising antibody response in Group II animals was significantly differentfrom the animals in Groups I and III. The response in Group I was the same as the control animals. Statistical significance (p value <0.05) was determined by non-parametricMann–Whitney test.

Treatment Animal Day

0 26 41 53 65 75 83

1st vaccination Boost 15 days after boost Challenge 12 days after challenge 22 days after challenge 30 days after challenge

Group I BTV CLPs

1 4 4 4 4 4 128 2562 4 4 4 4 32 64 2563 4 4 4 4 8 512 10244 4 4 4 4 32 256 2565 4 4 4 Died – – –6 4 4 4 4 256 512 10247 4 4 4 4 512 1024 1024

Mean 4.00 4.00 4.00 72.00 288.01 512SD 0.00 0.00 0.00 6.23 2.77 2.14

Mann–Whitney test with controls – p > 0.05 p > 0.05 p > 0.05 p > 0.05 p > 0.05Stat. signific. difference No differ No differ No differ No differ No differ No differ

Group II BTV-1 RSAVLPs

8 4 32 64 64 256 512 10249 4 32 64 64 1024 1024 102410 4 32 64 128 128 512 102411 4 32 64 128 128 256 102413 4 32 64 64 512 512 102414 4 32 128 64 512 1024 102415 4 16 32 128 1024 1024 1024

Mean 29.04 84.45 84.45 380.04 625.99 1024SD 1.30 1.41 1.41 1.40 1.69 0.00

Mann–Whitney test with controls p < 0.05 p < 0.05 p < 0.05 p < 0.05 p < 0.05 p < 0.05Stat. signific. difference Extremely Extremely Extremely Significant Significant Extremely

Group III salinea12 4 4 4 4 64 512 102416 4 4 4 4 256 512 51217 4 4 4 4 32 64 256

Saline and adjuvanta

18 4 4 4 4 256 512 51219 4 4 4 4 128 256 51220 4 4 4 4 128 256 51221 4 4 4 4 128 256 512

Mean 4.00 4.00 4.00 116.16 282.09 5120.00

e ther

a(

3v

dGifiamIrttisaHstK

r

lesions were observed at necropsy. Hydropericardium, petechialand ecchymotic haemorrhages in the heart muscle and pericardiumand discoloration of the myocardium were evident in two of the

SD 0.00 0.00

a Saline and saline + adjuvant treatments were taken as one control group becaus

fter challenge (day 65 onwards), with titres ranging from 4 to 512Table 1).

.4. Clinical response to virus challenge in VLP and CLPaccinated sheep

The clinical response and rectal temperature were monitoreduring the vaccination period and after challenge with BTV-1RE2007/01. No adverse responses, including any local reaction or

ncrease in rectal temperature were observed in the sheep after therst or second vaccination with the VLPs or CLPs or in the controlnimals vaccinated with saline or saline plus adjuvant. All ani-als were challenged with BTV-1 GRE2001/07. Animals in Group

I vaccinated with VLPs had a normal rectal temperature, whichemained consistent (∼38.9 ◦C) after challenge (Fig. 5). Only one ofhe CLP vaccinated animal (Group I), had elevated rectal tempera-ures (40.9 ◦C) at day 7 (Fig. 5). Consistent with the observation thatnfection of BTV infection of Greek sheep breeds is associated withubclinical BT symptoms, only 4 of 7 control animals (Group III) had

febrile period lasting 4 days that peaked 7–8 days after challenge.owever, none of the animals displayed any of the visible classical

igns of BT including nasal or ocular discharge, or lameness over

he duration of the experiment. This phenomena is typical for thearagouniko cross animals.

In order to investigate further the nature of the subclinicalesponse to virulent virus challenge, post-challenged animals at

2.09 2.09 1.49

e was no detectable statistical difference between them.

30 day were necropsied and major organs were examined for clas-sical BT haemorrhages. Although the clinical signs were mild orsubclinical in Groups I and III (CLP and control animals), distinct

Fig. 5. Temperature response of sheep challenged with virulent BTV-1. The rectaltemperatures (◦C) of BTV-1 VLP and BTV CLP vaccinated and saline vaccinated sheepwere monitored after challenge. None of the VLP vaccinated sheep had elevatedrectal temperature. The time of challenge is indicated by an arrow. The differentgroups are indicated; controls (�) BTV CLPs (©) and BTV-1 RSA VLPs (�).

M. Stewart et al. / Vaccine 30 (2012) 2131– 2139 2137

Fig. 6. Evaluation of viraemia in vaccinated animals following challenge with virulent BTV-1 Greece by RT-qPCR and ECE titration. Animals vaccinated with BTV-1 RSAVLP did not have detectable viraemia by RT-PCR or ECE. (A) Real-time RT-PCR quantification BTV-1. None of the animals had any detectable viral RNA prior to challenge.P detec( ntrols (�).

CFhpota

3r

ramdp(ieipwvrlmpsvtv

ost-challenge, none of the sheep vaccinated with BTV-1 VLPs vaccine (Group II) hadGroup III) had detectable viraemia. The average data for each group is indicated; coheep (Groups I and III). The different groups are indicated; control (�) and BTV CLP

LP-vaccinated animals (Group I) and 3 control animals (Group III).urthermore, one CLP vaccinated animal and one control animalad 1 cm wide haemorrhages in the tunica media at the base of theulmonary artery consistent with BTV infection. In contrast, nonef the animals vaccinated with VLPs in Group II showed any his-ological lesions consistent with BTV infection, indicating that allnimals were completely protected against the challenge virus.

.5. VLP vaccination in sheep prevented infectious challenge viruseplication

To determine whether VLPs and CLPs were able to inhibit theeplication of challenge virus in the animal, blood samples fromll animals including the control animals were collected and wereonitored for virus replication using two alternate methods as

escribed in Section 2. BTV was not detected in any of the sheeprior to challenge virus. None of the animals vaccinated with VLPsGroup II) had any detectable post-challenge viraemia when exam-ned by two alternate methods, real-time RT-PCR (Fig. 6A) andmbryonated chicken egg (ECE) inoculation (Fig. 6B). In contrast,nfectious and replicating virus was detectable in the blood sam-les of animals from Groups I and III. A typical viraemic responseas observed both by real-time RT-PCR and ECE inoculation. Peak

iraemia was observed at 8 days post-challenge, followed by a sloweduction of the virus in the circulating blood (Fig. 6). The viraload observed in CLP-vaccinated animals (Group I) was approxi-

ately 1 log10 lower than the control animals over the viraemiceriod (Fig. 6). Although these differences in viral load may not be

ignificant, it may indicate that CLPs may be involved in limitingirus spread. However, this data conclusively demonstrated thathe VLP-vaccinated animals were fully protected against virulentirus challenge and VLPs had prevented virus replication.

table viraemia. All animals vaccinated with BTV CLPs (Group I) and control animals (�) and BTV CLP (�). (B) BTV-1 Greece titration using ECEs samples collected from

4. Discussion

Traditionally, the control of BT disease in livestock relied on theuse of attenuated live virus vaccines and these vaccines have beenused routinely in Africa, where BTV is highly active. More recently,killed virus vaccines have been rapidly developed in response to theoutbreaks of BTV in European countries. Although both vaccineshave been demonstrated to be reasonably effective, they are stillassociated with certain inherent problems. Disadvantages of cur-rent attenuated vaccines for BTV include the inability to sufficientlyreduce post-vaccination viraemia from transmission [30], potentialreassortment and reversion of the attenuated vaccine strain [31]and side effects associated with vaccination [32]. Similarly, someof the concerns regarding current killed vaccines include potentialbatch-to-batch variation of virus inactivation and adverse reactionsin vaccinated animals [33] together with the inability to completelyinhibit viraemia in infected animals [34]. Most importantly, thesevaccines are not DIVA compliant and have a significant economicimpact due to restriction in animal trades and movements.

In contrast, the VLP vaccines assessed in this report demon-strated that VLP vaccinated animals were completely protectedfrom detectable post-challenge viraemia and had no side effects onanimal health. Furthermore, the VLP vaccines inhibited virus repli-cation, pyrexia and gross lesions in all post-challenged animals andafforded complete protective neutralising antibody response. Mostimportantly, the VLP vaccines also offered no threat to the healthof vaccinated animals, as there is no genetic material to revive thevirus. The VLP vaccine is also DIVA compliant due to the lack ofall four NS proteins and enzymatic proteins. Another advantage of

the VLP system is the ability to use the core elements to gener-ate vaccines to new serotypes by replacing VP2, the most variableprotein. VLP vaccines could reduce all regulatory hurdles quicklyand thus be ready for rapid rollout through the incorporation of

2 ccine 3

rckc

oiretssanploAtiprmtiratdvovoe

ie1plot(igtfcs(et

eBvotoG[v

btt

[

[

[

[

[

[

[

[

[

[

[

[

[

138 M. Stewart et al. / Va

elevant antigens, should an outbreak occur. Therefore, VLP vac-ines are not limited by concerns associated with attenuated andilled vaccines and, importantly, meet all requirements for a DIVAompliant vaccine.

Vaccination of animals with VP7 alone or in combination withther BTV proteins had been demonstrated to elicit a cell-mediatedmmune response and it is postulated that CMI response mayeduce BTV viraemia [35]. More recently, three CD8 and four CD4pitopes have been mapped within VP7 sequence [36]. To inves-igate this hypothesis in this report we vaccinated a group ofheep with CLPs consisting of VP7 and VP3, the two inner cap-id proteins. BTV CLPs induced a serological response against BTVs evidenced by the positive Ab-ELISA against VP7, but they didot induce neutralising antibody response. CLPs also failed torevent the signatures of BTV infection (e.g., pyrexia and gross

esions) in all post-challenged animals, and with the exceptionf one animal, none of the sheep had an elevated temperature.lthough the clinical signs of disease were mitigated, CLPs failed

o inhibit the replication of the challenge virus since high titres ofnfectious BTV-1 (>104 ECE50/ml) were detected in the blood sam-les of all vaccinated animals. The virus titres were only slightlyeduced in comparison to that of the control, non-vaccinated ani-als. The low level protection afforded by CLPs was probably due

o cell-mediated immune (CMI) response induced by VP7. Indeednduction of the CMI response in sheep after BTV infection had beeneported [37] and vaccination with VP7 might add to the protectionfforded by VP2 and VP5 [38]. Our results indicated VP7 in combina-ion with VP3 was not sufficient to significantly reduce viraemia orisease in infected animals and, therefore, would not be a suitableaccine candidate to prevent the spread of BTV via biting midgesr protect animals from disease. Based on these results, our pre-ious data [8,39,40] and others [41], it is clear that the presencef serotype specific neutralising antibodies in the BTV vaccines isssential for protection.

The incursion of BTV into the Mediterranean basin via Greeces considered to be one of the three major routes for the virus tonter Europe. This is highlighted by the outbreaks of BTV serotypes, 4, 9 and 16 via this route [42,43]. Furthermore, there is a distinctossibility that viruses from the same serotype with different evo-

utionary lineages will enter Europe via this route. The introductionf the same serotype from two different lineages is highlighted byhe emergence of BTV-1 into Europe via two separate routes in 1999via Greece) and 2006 (via Iberian Peninsula). In this report, we havenvestigated the importance of the diverse lineages within a sin-le serotype for vaccine development. Indeed, some would suggesthat the evolutionary lineage of BTV should be taken into accountor designing an effective BTV vaccine. Our data here demonstratedlearly that all animals vaccinated with VLPs developed serotypepecific neutralising antibodies against not only the vaccine strainwestern lineage) but also the challenge virus strain (eastern lin-age) suggesting that isolate variation or evolutionary distance ofhe strain is not a critical factor for BTV vaccine.

Our data also showed that BTV-1 VLP vaccine of a western lin-age completely inhibited virus replication of the eastern lineageTV-1 GRE2001/07 isolate. This is an additional advantages of VLPaccine and directly relevant to limiting the possible transmissionf the virus. This is particularly relevant for BTV infection in cer-ain breeds of animals, which do not exhibit any clinical disease ornly subclinical signs (e.g., Karagouniko and Lesbos sheep breeds ofreece) although virus could replicate very efficiently to high titres

2]. Thus, these animals are the perfect reservoirs for infectiousirus replication and thereby suitable for dispersal by midges.

Our results demonstrated the development of neutralising anti-odies are critical for complete protection and strongly suggesthat evolutionary lineage of a BTV serotype is not a critical fac-or and does not have to be taken into account when rationally

[

0 (2012) 2131– 2139

designing a candidate vaccine. This is particularly important forEurope as outbreaks of viruses from the same serotype but differ-ent geographical lineage frequently occur. The role of evolutionarylineage is only important as an epidemiological tool.

Acknowledgements

This study was funded by EU (FP6-2005-SSP-5A). The authorswould like to thank Dr. V. Psychas (Aristotle University of Thessa-loniki, Greece) for assessing gross lesions in necropsied animals.

References

[1] Roy P. Orbiviruses. In: Knipe DM, Howley PM, editors. Fields’ Virology. 5th ed.Philadelphia, NY, USA: Lippincott Williams & Wilkins; 2007. p. 1975–97.

[2] Koumbati M, Mangana O, Nomikou K, Mellor PS, Papadopoulos O. Durationof bluetongue viraemia and serological responses in experimentally infectedEuropean breeds of sheep and goats. Vet Microbiol 1999;64(4):277–85.

[3] Nomikou K, Mangana-Vougiouka O, Panagiotatos DE. Overview of bluetonguein Greece. Vet Ital 2004;40(3):108–15.

[4] Maan S, Maan NS, Nomikou KSJ, Anthony SN, Ross-Smith N, Singh KAR, et al.,editors. Molecular epidemiology studies of bluetongue virus. Academic Press;2009.

[5] Savini G, Hamers C, Conte A, Migliaccio P, Bonfini B, Teodori L, et al. Assessmentof efficacy of a bivalent BTV-2 and BTV-4 inactivated vaccine by vaccination andchallenge in cattle. Vet Microbiol 2008.

[6] Boone JD, Balasuriya UB, Karaca K, Audonnet JC, Yao J, He L, et al. Recombinantcanarypox virus vaccine co-expressing genes encoding the VP2 and VP5 outercapsid proteins of bluetongue virus induces high level protection in sheep.Vaccine 2007;25(4):672–8.

[7] Perrin A, Albina E, Breard E, Sailleau C, Prome S, Grillet C, et al.Recombinant capripoxviruses expressing proteins of bluetongue virus: eval-uation of immune responses and protection in small ruminants. Vaccine2007;25(37–38):6774–83.

[8] Stewart M, Bhatia Y, Athmaram TN, Noad R, Gastaldi C, Dubois E, et al. Valida-tion of a novel approach for the rapid production of immunogenic virus-likeparticles for bluetongue virus. Vaccine 2010;28(17):3047–54.

[9] Roy P. Use of novel technology for dissecting of Bluetongue virus assembly anddevelopment of candidate vaccine. In: International symposium on virus cellinteraction: cellular and molecular responses. 1994. p. 91–102.

10] Roy P. Development of Bluetongue virus subunit vaccines. In: The proceedingsof Arbovirus meeting. 1990. p. 328–32.

11] Huismans H. The use of recombinant DNA technology for the development of abluetongue virus subunit vaccine. Onderstepoort J Vet Res 1985;52(3):149–51.

12] Perez de Diego AC, Athmaram TN, Stewart M, Rodriguez-Sanchez B, Sanchez-Vizcaino JM, Noad R, et al. Characterization of protection afforded by a bivalentvirus-like particle vaccine against bluetongue virus serotypes 1 and 4 in sheep.PLoS One 2011;6(10):e26666.

13] Roy P, Bishop DHL, LeBlois H, Erasmus BJ. Long-lasting protection ofsheep against bluetongue challenge after vaccination with virus-like parti-cles: evidence for homologous and partial heterologous protection. Vaccine1994;12(9):805–11.

14] Nomikou K, Dovas CI, Maan S, Anthony SJ, Samuel AR, Papanastassopoulou M,et al. Evolution and phylogenetic analysis of full-length VP3 genes of EasternMediterranean bluetongue virus isolates. PLoS One 2009;4(7):e6437.

15] French TJ, Roy P. Synthesis of bluetongue virus (BTV) corelike particles by arecombinant baculovirus expressing the two major structural core proteins ofBTV. J Virol 1990;64(4):1530–6.

16] Noad RJ, Stewart M, Boyce M, Celma CC, Willison KR, Roy P. Multigene expres-sion of protein complexes by iterative modification of genomic Bacmid DNA.BMC Mol Biol 2009;10:87.

17] Belyaev AS, Roy P. Development of baculovirus triple and quadruple expres-sion vectors: co-expression of three or four bluetongue virus proteins and thesynthesis of bluetongue virus-like particles in insect cells. Nucleic Acids Res1993;21(5):1219–23.

18] French TJ, Marshall JJ, Roy P. Assembly of double-shelled, virus-like particles ofbluetongue virus by the simultaneous expression of four structural proteins. JVirol 1990;64(12):5695–700.

19] Marshall JJ, Roy P. High level expression of the two outer capsid proteins ofbluetongue virus serotype 10: their relationship with the neutralization of virusinfection. Virus Res 1990;15(3):189–95.

20] van der Walt NT. A haemagglutination and haemagglutination inhibition testfor bluetongue virus. Onderstepoort J Vet Res 1980;47(2):113–7.

21] Chatzinasiou E, Dovas CI, Papanastassopoulou M, Georgiadis M, Psychas V,Bouzalas I, et al. Assessment of bluetongue viraemia in sheep by real-time PCRand correlation with viral infectivity. J Virol Methods 2010;169(2):305–15.

22] Breard E, Sailleau C, Coupier H, Mure-Ravaud K, Hammoumi S, Gicquel B, et al.

Comparison of genome segments 2, 7 and 10 of bluetongue viruses serotype2 for differentiation between field isolates and the vaccine strain. Vet Res2003;34(6):777–89.

23] Clavijo A, Heckert RA, Dulac GC, Afshar A. Isolation and identification of blue-tongue virus. J Virol Methods 2000;87(1–2):13–23.

cine 3

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

M. Stewart et al. / Vac

24] Goldsmit L, Barzilai E. An improved method for the isolation and identificationof bluetongue virus by intravenous inoculation of embryonating chicken eggs.J Comp Pathol 1968;78(4):477–87.

25] Boyce M, Roy P. Recovery of infectious bluetongue virus from RNA. J Virol2007;81(5):2179–86.

26] Maan S, Maan NS, Samuel AR, Rao S, Attoui H, Mertens PP. Analysis and phy-logenetic comparisons of full-length VP2 genes of the 24 bluetongue virusserotypes. J Gen Virol 2007;88(Pt 2):621–30.

27] Siegel S, Castellan N. Non parametric statistic for the behavioral sciences. 2nded. New York: McGraw-Hill Book Co.; 1988.

28] DeMaula CD, Heidner HW, Rossitto PV, Pierce CM, MacLachlan NJ. Neutrali-sation determinants of United States bluetongue virus serotype 10. Virology1993;195(1):292–6.

29] DeMaula CD, Bonneau KR, MacLachlan NJ. Changes in the outer capsid proteinsof bluetongue virus serotype ten that abrogate neutralisation by monoclonalantibodies. Virus Res 2000;67(1):59–66.

30] Veronesi E, Hamblin C, Mellor PS. Live attenuated bluetongue vaccine virusesin Dorset Poll sheep, before and after passage in vector midges (Diptera: Cer-atopogonidae). Vaccine 2005;23(48–49):5509–16.

31] Batten CA, Maan S, Shaw AE, Maan NS, Mertens PP. A European field strain ofbluetongue virus derived from two parental vaccine strains by genome segmentreassortment. Virus Res 2008;137(1):56–63.

32] Savini G, Cannas A, Casaccia C, Di Gialleonardo L, Leone A, Patta C, et al. Risk fac-tors associated with the occurrence of undesired effects in sheep and goats afterfield vaccination with modified-live vaccine against bluetongue virus serotypes2, 4 and 16. Vet Microbiol 2010;146(1–2):44–50.

33] Gonzalez JM, Figueras L, Ortega ME, Lozano M, Ruiz de Arcaute M, Royo R,et al. Possible adverse reactions in sheep after vaccination with inactivatedBTV vaccines. Vet Rec 2010;166:757–8.

34] van der Sluijs MT, Schroer-Joosten DP, Fid-Fourkour A, Vrijenhoek MP, DebyserI, Gregg DA, et al. Effect of vaccination with an inactivated vaccine on

[

0 (2012) 2131– 2139 2139

transplacental transmission of BTV-8 in mid term pregnant ewes and heifers.Vaccine 2012;30(3):647–55.

35] Jeggo MH, Wardley RC, Brownlie J. Importance of ovine cytotoxic T cells in pro-tection against bluetongue virus infection. Prog Clin Biol Res 1985;178:477–87.

36] Rojas JM, Rodriguez-Calvo T, Pena L, Sevilla N. T cell responses to bluetonguevirus are directed against multiple and identical CD4+ and CD8+ T cell epitopesfrom the VP7 core protein in mouse and sheep. Vaccine 2011;29(40):6848–57.

37] Jeggo MH, Wardley RC. Generation of cross-reactive cytotoxic T lympho-cytes following immunization of mice with various bluetongue virus types.Immunology 1982;45(4):629–35.

38] Calvo-Pinilla E, Rodriguez-Calvo T, Sevilla N, Ortego J. Heterologous primeboost vaccination with DNA and recombinant modified vaccinia virusAnkara protects IFNAR(−/−) mice against lethal bluetongue infection. Vaccine2009;28(2):437–45.

39] Roy P, Bishop DH, LeBlois H, Erasmus BJ. Long-lasting protection of sheepagainst bluetongue challenge after vaccination with virus-like particles:evidence for homologous and partial heterologous protection. Vaccine1994;12(9):805–11.

40] Huismans H, Van der Walt NT, Cloete M, Erasmus BJ. Isolation of a capsid pro-tein of bluetongue virus that induces a protective immune response in sheep.Virology 1987;157(1):172–9.

41] Wade Evans AM, Romero CH, Mellor P, Takamatsu H, Anderson J, ThevasagayamJ, et al. Expression of the major core structural protein (VP7) of bluetongue virus,by a recombinant capripox virus, provides partial protection of sheep against avirulent heterotypic bluetongue virus challenge. Virology 1996;220(1):227–31.

42] Purse BV, Mellor PS, Rogers DJ, Samuel AR, Baylis CMPPM. Climate change

and the recent emergence of bluetongue in Europe. Nat Rev Microbiol2005;3:171–81.

43] Mellor PS, Carpenter S, Harrup L, Baylis M, Mertens PP. Bluetongue in Europeand the Mediterranean Basin: history of occurrence prior to 2006. Prev Vet Med2008;87(1–2):4–20.