Evp

SMAa

b

a

ARR2AA

KFSRCC

1

enstTcapava

0d

Vaccine 29 (2011) 1906–1912

Contents lists available at ScienceDirect

Vaccine

journa l homepage: www.e lsev ier .com/ locate /vacc ine

valuation of cross-protection between O1 Manisa and O1 Campos in cattleaccinated with foot-and-mouth disease virus vaccine incorporating differentayloads of inactivated O1 Manisa antigen

inganallur Balasubramanian Nagendrakumara, Villuppanoor Alwar Srinivasana,∗,uthukrishnan Madhanmohana, Shanmugam Yuvaraja, Satya Paridab,

ntonello Di Nardob, Jacquelyn Horsingtonb, David James Patonb

Foot-and-Mouth Disease Virus Laboratory, Research and Development Centre, Indian Immunologicals Limited, Rakshapuram, Gachibowli Post, Hyderabad 500032, IndiaPirbright Laboratory, Institute for Animal Health, Ash Road, Woking, Surrey GU240NF, United Kingdom

r t i c l e i n f o

rticle history:eceived 21 June 2010eceived in revised form2 December 2010ccepted 24 December 2010vailable online 13 January 2011

eywords:MDVerologyelative homology

a b s t r a c t

Serology is used to predict vaccine induced protection against challenge with a heterologous strain of thesame serotype of foot-and-mouth disease virus (FMDV). To evaluate the accuracy of such predictions,we compared the protection afforded to cattle vaccinated with the O1 Manisa strain of FMDV againstchallenge with either a homologous (O1 Manisa) or a heterologous strain (O1 Campos). Serology by virusneutralization test (VNT) using O1 Manisa antiserum predicted an acceptable protection against such achallenge. Two experiments were carried out to compare the results for consistency. A total of 78 naïvecattle were vaccinated with different antigen payloads (60–0.94 �g) of O1 Manisa. They were challengedby intradermolingual inoculation with live FMDV, either O1 Manisa or O1 Campos. Unvaccinated naïvecontrol cattle (n = 20) were also challenged with either the O1 Manisa or O1 Campos viruses and all devel-oped generalized FMD. The protection results for the vaccinated cattle revealed that higher payloads of O1

ross-protectionarrier status

Manisa vaccine were needed to protect against heterologous challenge compared to that for homologouschallenge. The 50% protective dose (PD50) values for the vaccine in experiments 1 and 2 were found tobe 28.78 and 9.44 for the homologous challenge and 3.98 and 5.01 for heterologous challenge. Further-more, protection against O1 Campos required a higher level of vaccine-induced antibody against this viruscompared to the level of O1 Manisa neutralizing antibody associated with protection against homologouschallenge. The 50% protective level of in vitro neutralizing antibody was found to be log10 1.827 for O1

or O1

Campos and log10 0.954 f

. Introduction

Foot-and-mouth disease virus (FMDV) exists as seven differ-nt serotypes and infection or vaccination with one serotype doesot protect against the others [1,2]. In addition, many antigenictrains have been recognized within serotypes [3,4] and some ofhese differences may be important in relation to cross-protection.herefore, serological tests are routinely used as part of the pro-ess for selecting the most appropriate vaccine strain for protectiongainst a given field isolate [5,6]. The mechanisms of the immune

rotection elicited by vaccination are not fully understood [7,8]nd relatively few published reports confirming the predictivealue of serological vaccine matching tests [9,10,11,12] are avail-ble. Therefore, a study was undertaken to evaluate the accuracy

∗ Corresponding author. Tel.: +91 40 23000894; fax: +91 40 23005958.E-mail address: srini@indimmune.com (V.A. Srinivasan).

264-410X/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.oi:10.1016/j.vaccine.2010.12.127

Manisa based on O1 Manisa based virus neutralization test.© 2011 Elsevier Ltd. All rights reserved.

of serologically predicted cross-protection within a serotype byvaccinating cattle with the O1 Manisa strain of FMDV and thenchallenging them with either a homologous (O1 Manisa) or a het-erologous strain (O1 Campos).

2. Materials and methods

2.1. Cell lines and viruses

Baby Hamster kidney (BHK) and primary bovine thyroid (BTY)cells were provided by the tissue culture laboratory of IndianImmunologicals Limited (IIL), Hyderabad. The O1 Manisa strain ofFMD virus is derived from a 1968 Turkish isolate that is of the

Middle-East South-Asian topotype, whereas the O1 Campos strainwas originally isolated in Brazil in 1958 and is of the European SouthAmerican topotype [13]. The O1 Manisa vaccine strain obtainedfrom the World Reference Laboratory, Institute for Animal Health(WRL-IAH), Pirbright, United Kingdom, was maintained in the virus

S.B. Nagendrakumar et al. / Vaccine 29 (2011) 1906–1912 1907

Table 1Experiment groups and results of percentage protection against homologous and heterologous challenge on 21 dpv in cattle calves vaccinated with different payloads of O1

Manisa vaccine.

Groups Payload (�g) Challenge virus Results of challenge

Experiment 1 Experiment 2

Numberchallenged

Numberprotected

Percentageprotection

Numberchallenged

Numberprotected

Percentageprotection

Group 1 60 O1 Campos 8 6 75 5 5 100Group 2 15 O1 Manisa 8 8 100 5 5 100Group 3 15 O1 Campos 8 4 50 5 3 60Group 4 3.75 O1 Manisa 8 8 100 5 5 100

scdaoMu2tvaLpatlwlFCL

a1e1lptua

2

uivwH

2

oTapLa

Group 5 3.75 O1 Campos 7 2Group 6 0.94 O1 Manisa 8 7Group 7 UV control O1 Campos 8 0Group 8 UV control O1 Manisa 8 0

eed laboratory, IIL, Hyderabad, as a BHK-CZ adapted suspensionulture for preparation of antigen used to formulate vaccines withifferent payloads. Oil adjuvant vaccines incorporating differentntigen payloads determined by spectrophotometric quantitationf 146S antigen [14] were blended as double oil emulsion usingontanide Oil ISA 206 (Seppic, France). Different antigen lots were

sed for formulating the vaccine for experiment 1 and experiment. O1 Campos vaccine strain received from WRL-IAH was main-ained in the FMDV Laboratory, IIL, Hyderabad. Cattle challengeiruses O1 Manisa and O1 Campos were prepared from the BHK-21dapted virus obtained from the virus seed laboratory and FMDVaboratory, IIL, Hyderabad respectively. These viruses had beenassaged and titrated in cattle tongue. The titrated viruses wereliquoted and stored as 50% glycerol stock and the same batch ofhe challenge virus was used for both the experiments. For chal-enge experiments 10,000 bovine tongue infectious dose (BTID50)

ere used for the intradermolingual challenge of each calf. Fol-owing challenge, the challenge virus was titrated in BTY cells.or serum neutralization assays virus pools of O1 Manisa and O1ampos maintained at the Institute for Animal Health, Pirbrightaboratory, were used.

Four different monovalent vaccine formulations of fixed totalnd adjuvant volumes were prepared using O1 Manisa virus (60,5, 3.75 and 0.94 �g) in Montanide Oil ISA 206 (Seppic, France). Forxperimental consideration vaccine formulations with payloads 60,5 and 3.75 �g would simulate a potency test for heterologous chal-

enge while formulations with 15, 3.75 and 0.94 would simulate aotency test for homologous challenge. This is a modification fromhe usual potency test as prescribed by OIE where reduced dose vol-me of a vaccine formulation is administered as neat, one fourthnd one sixteenth dilution.

.2. Serological cross-reactivity

Bovine vaccinate serum (BVS) prepared at IIL and at WRL-IAHsing O1 Manisa monovalent vaccine was used to measure the

n vitro relative homology (r1) value of O1 Campos to O1 Manisaaccine strain. Two dimensional micro neutralization assay (MNT)as performed as per the method described by Rweyemamu andingley [15].

.3. Cattle immunization and challenge

Ninety-eight Jersey or Friesian cross bred cattle (10–12 monthsf age) were obtained from the holding farm of IIL, Hyderabad.

hese animals were reared in the farm from one month of agend were screened by 3 rounds of testing for FMDV-non-structuralrotein (NSP) antibodies using PrioCHECK® FMDV NS kit (Prionicselystad B.V., The Netherlands). All the animals were NSP seroneg-tive in all the three tests. Additionally, the animals were tested for

28.6 5 0 087.5 5 1 20

0 2 0 00 2 0 0

the absence of virus in the oesophago-pharyngeal fluids (probangsamples) thrice by virus isolation on primary bovine thyroid cells[16] followed by antigen ELISA and RT-PCR [17].

For experiment 1, sixty-four animals were used (vaccinatedn = 48; unvaccinated n = 16) while for experiment 2, thirty-fouranimals were used (Vaccinated n = 30; Unvaccinated n = 4). The ani-mals were observed for appearance of clinical signs of FMD and therectal temperature was recorded daily for 10 days post challenge.Probang and blood samples were taken on 0 (day 21 post vaccina-tion; 21 dpv), 5, 10, 15, 21 and 35 days post-challenge (dpc) to testfor virus genome and antibodies respectively.

2.3.1. Experiment 1Three groups of eight animals each (Groups 1, 3, 5) were vac-

cinated with FMDV monovalent vaccine incorporating 60, 15 and3.75 �g of O1 Manisa antigen respectively. Three groups of eightanimals each (Groups 2, 4, 6) were vaccinated with FMDV mono-valent vaccine incorporating 15, 3.75 and 0.94 �g of O1 Manisaantigen respectively. Groups 1, 3 and 5 were challenged by intra-dermolingual route with O1 Campos virus (heterologous challenge)and Groups 2, 4 and 6 were challenged by intradermolingualroute with O1 Manisa virus (homologous challenge) on day 21post-vaccination respectively (Table 1). Two groups of eight unvac-cinated control cattle (Group 7 and Group 8) were also challengedwith O1 Campos and O1 Manisa respectively.

2.3.2. Experiment 2A second experiment with 34 male cattle calves was performed

with 5 animals in each of three groups for homologous challengewith O1 Manisa and 5 animals in each of three groups for heterol-ogous challenge with O1 Campos, while 2 animals each were keptas unvaccinated controls (Table 1) and challenged on 21 dpv. Thevaccine groups and challenge groups were similar to experiment 1.

2.4. Real-time quantitative RT-PCR assay for detection of viralRNA

For automated RNA extraction from oesophago-pharyngealfluid (probang samples), 200 �l of sample was added to 300 �lMagNA Pure LC total nucleic acid lysis buffer (Roche, Welwyn Gar-den City, UK). Total nucleic acids were extracted and eluted in 50 mlelution buffer by using the MagNA Pure LC total nucleic acid iso-lation kit (Roche) with an automated robotic workstation (Roche,Switzerland), according to manufacturer’s instructions. ExtractedRNA was stored at −80 ◦C until used.

The amount of viral RNA was quantified by two-step real-timeRT-PCR (rRT-PCR) as described previously [18], using a universalTaqMan probe specific for FMDV [19]. PCR assays were performedin a Stratagene MX4000 machine. For the generation of standardcurves, FMDV RNA standard was synthesized in vitro from a plasmid

1 / Vacc

csK(

2

wnr1

2

Sta((ov3TcaDRmpmaTeewptt<gwsrs

2

nPBs

3

3

Caa

908 S.B. Nagendrakumar et al.

ontaining a 500 base pair insert of the internal ribosomal entryite (IRES) of FMDV O UKG 34/2001 (kindly provided by Dr. Donalding, Institute for Animal Health, UK) using MEGAScripTM T7 kit

Ambion, USA) [20].

.5. Serology for virus neutralizing antibodies

Serum samples collected on 0 (21 dpv), 5, 10, 15, 21 and 35 dpcere examined for anti-FMDV neutralizing antibodies [21]. Theeutralizing antibody titres were calculated as the log10 of theeciprocal antibody dilution required for 50% neutralization of00 TCID50 virus.

.6. Quantitative analyses

The 50% potency dose (PD50) was estimated by thepearman–Karber method [22]. The dataset used to constructhe model comprised 78 individual serum antibody titres takent 21 dpv from 13 animals vaccinated with the 15 �g of antigenexperiments 1 and 2), 13 animals receiving 3.75 �g of antigenexperiments 1 and 2) and 13 animals vaccinated with 0.94 �gf antigen (experiments 1 and 2) and challenged with O1 Manisairus and a similar number of animals vaccinated with 60, 15,.75 �g of antigen respectively challenged with O1 Campos virus.he 20 unvaccinated control animals (10 each for O1 Manisahallenge and O1 Campos challenge from both experiments 1nd 2) were also included for the purpose of model building.escriptive statistical analyses were performed and the Logisticegression model was parameterised using the glm (binomialodel, link = logit) command in R 2.11.1 (www.R-project.org) for

redicting the PA50 values [23]. Protection was treated as dichoto-ous variable (1 = protected against challenge; 0 = not protected

gainst challenge) and entered in the model as dependent term.he 50% endpoint of serum neutralizing antibody titres at 21 dpvxpressed as log10 was entered as independent term. Predictedstimates from the GLM model were used to return the PA50 valuesith 95% confidence interval returned by R via confint() commandrovided by the package MASS. Graphs were constructed usinghe ggplot2 package for R. Since the serum neutralization antibodyitres do not follow normal distribution and the sample size is30, the 95% confidence intervals were calculated for each vaccineroup based on the t-distribution assumption (x ± t˛/2s/

√n),

here: x is the mean; s is the sample standard deviation; n is theample size; t˛/2 is the t value of an area of ˛/2 to its right [24]. Theesults of rRT-PCR on the virus excretion detected in the probangamples were analyzed statistically by standard methods.

.7. Virus identities established by sequencing

The common identity of the vaccine, challenge and in vitroeutralization test viruses of the same name was checked by1 sequence comparison using modified Sanger method usingigDye® Termination kit (ABI Prism, USA) in an automatedequencer (ABI Prism, USA).

. Results

.1. Serological cross-reactivity

The relative homology value (r1) between O1 Manisa and O1ampos derived using two dimensional micro neutralization testnd previously established bovine vaccinate sera was 0.64 and 0.62t Hyderabad and Pirbright respectively.

ine 29 (2011) 1906–1912

3.2. Serology for neutralizing antibody response

3.2.1. Experiment 1At 21 days post-vaccination (dpv), most of the vaccinated cat-

tle had sero-converted to O1 Manisa with neutralizing antibodytitres averaging 2.19, 1.67, 1.56, 1.52, 1.82 and 1.45 log10 SN50 forGroups 1–6 respectively (Fig. 1). Sera were also checked for theirlevels of neutralizing antibody to the heterologous virus O1 Cam-pos (2.33, 1.64, 1.60, 1.71, 2.05 and 1.73 log10 SN50 for Groups 1–6respectively) (Fig. 1) which were slightly higher (P < 0.05; t = −2.46;P = 0.018 based on paired t-test) when compared with the O1 Man-isa titres. Six animals (#761 in Group 2, #973 in Group 3, #272 and273 in Group 4 and #337 and 340 in Group 6) did not respond tovaccination and had poor antibody titres (log10 SN50 < 0.61) on day21 post vaccination (i.e. on the day of challenge).

3.2.2. Experiment 2At 21 days post-vaccination (dpv), most of the vaccinated cattle

had sero-converted to O1 Manisa with neutralizing antibody titresaveraging 1.96, 1.93, 1.96, 1.81, 1.36 and 1.36 log10 SN50 for Groups1–6 respectively (Fig. 1). Sera were also checked for their levels ofneutralizing antibody to the heterologous virus O1 Campos (1.69,1.33, 1.51, 0.97, 1.06 and 0.91 log10 SN50 for Groups 1–6 respec-tively) (Fig. 1) which were lower (P < 0.01; t = 7.00; P = 0.000 basedon paired t-test) when compared with the O1 Manisa titres. All butone animal (#4092 in Group 5) responded to vaccination.

3.3. Cattle protection experiments

3.3.1. Experiment 1The results of cattle protection experiments are given in Table 1.

All vaccinated cattle (n = 8) in Groups 2 and 4 and seven out of eightcattle in Group 6 were protected against FMD after challenge withO1 Manisa virus, whereas six out of eight in Group 1, two out of theeight cattle in Group 3, two out of seven in Group 5 (one animaldied on 4 dpc) were protected against challenge with O1 Camposvirus. All control animals in Groups 7 and 8 developed generalizedFMD, although delayed onset of clinical signs was noticed in onecontrol animal (#1001) challenged with O1 Manisa.

3.3.2. Experiment 2The results of cattle protection experiments are given in Table 1.

All five O1 Manisa vaccinated and challenged cattle in Groups 2and 4 were protected. Only one out of five cattle in Group 6 wasprotected. In the groups that were vaccinated with O1 Manisa andchallenged with O1 Campos all five in Group 1, three out of five inGroup 3 and none of the animals in Group 5 were protected. Allcontrol animals in Groups 7 and 8 developed generalized FMD.

3.4. Quantitation of FMDV RNA by rRT-PCR from probang samples

The results are summarised as mean copy numbers per groupin Fig. 2. In Manisa challenged cattle, more viral RNA was shed byanimals in the unvaccinated group and in the lowest dose vaccinegroup (Group 6, 0.94 �g), with differences being significant fromboth the other groups (Groups 2 and 4; 15 �g and 0.94 �g) at 5 dpc.In Campos challenged cattle, the highest shedding was also in theunvaccinated Group 7 and in the lowest dose Group 5 (3.75 �g) at

5 dpc. The levels of shedding were similar in the unvaccinated andlowest dose vaccine groups at 5 dpc for Manisa and Campos chal-lenged animals. Other groups tended to shed more Campos viralgenome than Manisa viral genome at equivalent dosages and timepoints.

S.B. Nagendrakumar et al. / Vaccine 29 (2011) 1906–1912 1909

with

3p

ccOwri0T

3

waf(seAms

4

c

Fig. 1. Serum antibody response towards O1 Manisa vaccine in animals challenged

.5. Estimation of protective antigen dose50 (PD50) and ofrotective antibody titre50 (PA50)

The PD50 values for experiments 1 and 2 for homologoushallenge were 28.78 and 9.44 respectively and for heterologoushallenge were 3.98 and 5.01 respectively. The predicted PA50 for1 Manisa with O1 Manisa VNT results and O1 Campos VNT resultsere 0.954 (95% CI 0.194–3.533) and 0.797 (95% CI 0.121–3.506)

espectively while the predicted PA50 for O1 Campos with O1 Man-sa VNT results and O1 Campos VNT results were 1.827 (95% CI.440–7.715) and 1.798 (95% CI 0.441–7.538) respectively (Fig. 3;able 3).

.6. Virus identities established by sequencing

The complete nucleotide sequences of the P1 encoding genesere determined for the vaccine and each challenge virus as well

s the viruses used for in vitro serology. Virus isolated from theoot lesion collected on 5 dpc of an O1 Manisa challenged animal#3930) and an O1 Campos challenged animal (#4167), that hadufficient antibody titres against homologous antigen and werexpected to be protected were characterized by P1 sequencing.comparison of the deduced amino acid sequences revealed noajor changes in the antigenic sites of the vaccine strain, challenge

train and the isolated virus (Table 2).

. Discussion

Prior to the start of this cross-challenge experiment, serologicalross-reactivity between O1 Manisa and O1 Campos was examined

(a) O1 Manisa and (b) O1 Campos carried out as two separate sets of experiments.

independently at Hyderabad and Pirbright using separate stocks ofviruses and antisera. The ability of bovine post vaccination sera toneutralize or bind to O1 Campos virus was compared with the reac-tivity of the same sera in the same tests against parental O1 Manisavirus. Serology by virus neutralization test (VNT) using O1 Man-isa antiserum predicted an acceptable protection (r1 = 0.64/0.62 atHyderabad and Pirbright respectively, where >0.3 indicates that thefield isolate is sufficiently similar to the vaccine strain that use of thevaccine is likely to confer protection against challenge with the fieldisolate) against such a challenge. The OIE recommended approachfor determining the 50% potency dose (PD50) involves administra-tion of different volumes of the vaccine to each group of cattle [25].For example, if the label states 2 ml as the full dose, a 1/4 and a 1/16dose of vaccine can be obtained by injecting 0.5 ml and 0.125 mlrespectively. This method preserves a constant ratio of antigen toadjuvant. In the present experiment, the vaccine dose remainedconstant while different vaccine formulations incorporating vari-ous antigen payloads representing the fourfold dilutions (1/4 and1/16) were used. Such vaccine formulations have been reportedearlier [26]. The diluents used in these methods were immunolog-ically active. The advantage of such a method is that all the animalsget the same volume of vaccine but with different payloads andadministration of vaccine with the fourfold dilution is not difficult.However, it has been argued that such methods offer lower probitslope and result in an increased heterogeneity of the testing system.

For determining the protective index for homologous challengewith O1 Manisa virus, three groups of cattle (Groups 2, 4 and 6)were vaccinated with 15, 3.75 and 0.94 �g O1 Manisa antigen andfor determining the protective index for heterologous challengewith O1 Campos virus, three groups of cattle (Groups 1, 3 and 5)

1910 S.B. Nagendrakumar et al. / Vaccine 29 (2011) 1906–1912

Fig. 2. Mean of copy numbers of FMDV detected at 5, 10, 15, 21 and 35 days post-challenge by quantitative RT-PCR with different Ag payloads used for homologous (a) andheterologous (b) challenges. Error bars represent 95% CI.

Fig. 3. Scatter plot of (a) O1 Manisa and (b) O1 Campos serological responses at 21 days post-vaccination and proportion protected against homologous and heterologouschallenges with fitted lines for the Logistic Regression model and corresponding 95% CIs. Predicted PA50 values for O1 Manisa and O1 Campos viruses in O1 Manisa vaccinatedcattle. Data from both experiment 1 and 2 were included to plot the graphs. PA50 for O1 Manisa with O1 Manisa VNT results: 0.954; PA50 for O1 Campos with O1 Manisa VNTresults: 1.827; PA50 for O1 Manisa with O1 Campos VNT results: 0.797; PA50 for O1 Campos with O1 Campos VNT results: 1.798.

S.B. Nagendrakumar et al. / Vaccine 29 (2011) 1906–1912 1911

Table 2Comparison of sequence identities of the viruses used in vaccine production, cattle challenge test and neutralization test.

Virus strain Virus origin VP4 VP3 VP1

O1 Manisa O1 Manisa69 [GenBank ID: AJ251477] aa9-T, aa40-D, aa61-N aa139-A, aa-163S aa57-P, aa137-GBHK7 (virus used for serology) aa9-P, aa40-D, aa61-N aa139-A, aa-163S aa57-R, aa137-VBHKCZ (vaccine strain) aa9-P, aa40-D, aa61-N aa139-A, aa-163S aa57-R, aa137-VCattle passage 1 aa9-P, aa40-E, aa61-K aa139-P, aa-163P aa57-P, aa137-VCattle passage 1 aa9-P, aa40-E, aa61-K aa139-P, aa-163P aa57-R, aa137-VCattle challenge virus aa9-P, aa40-E, aa61-K aa139-P, aa-163P aa57-P, aa137-VAffected cattle – 3930 foot epithelium aa9-P, aa40-E, aa61-K aa139-P, aa-163S aa57-P, aa137-V

O1 Campos O1 Campos58 [GenBank ID: AJ320488] No change No change aa4-A, aa13-T, aa-97-G, aa156-VBHK5 No change No change aa4-A, aa13-T, aa-97-G, aa156-VBHK6 (Virus used for serology) No change No change aa4-A, aa13-T, aa-97-G, aa156-VCattle passage 1 No change No change aa4-V, aa13-A, aa-97-A, aa156-ACattle passage 1 No change No change aa4-V, aa13-A, aa-97-A, aa156-ACattle challenge virus No change No change aa4-V, aa13-A, aa-97-A, aa156-AAffected cattle – 4167 foot epithelium No change No change aa4-A, aa13-A, aa-97-A, aa156-A

No mutations in VP2 region of both viruses were noticed.

Table 3Evaluation of O1 Manisa and O1 Campos SNTs in the Logistic Regression model for predicting the PA50 values by homologous and heterologous challenge.

Challenge virus ˇ SE (ˇ) 95% CI Z P

O1 Manisa Constant −2.766 0.969 −4.664 to −0.867 −2.86 0.000O1 Manisa VNT (log10 SN50 titres) 2.898 0.805 1.320 to 4.476 3.60 0.004Constant −2.310 0.893 −4.060 to −0.561 −2.59 0.010O1 Campos VNT (log10 SN50 titres) 2.898 0.888 1.158 to 4.638 3.26 0.001

O1 Campos Constant −4.183 1.302 −6.735 to −1.631 −3.21 0.001

wuoiO(gc

ehs(icsh

tiftatcl3sbd7cprta

O1 Manisa VNT (log10 SN50 titres) 2.289Constant −3.545O1 Campos VNT (log10 SN50 titres) 1.971

ere vaccinated with 3.75, 15 and 60 �g O1 Manisa antigen. Thenvaccinated control animals challenged with either O1 Manisar O1 Campos showed generalization of disease and typical clin-cal signs of FMD except in the case of one control animal in the

1 Manisa challenge group that showed delayed generalization9 dpc). The PD50 and PA50 results contrast with the prediction of aood match between O1 Manisa and O1 Campos based on in vitroross-neutralization.

In the case of serotype A FMDV vaccines, a recent study by Brehmt al. [12] evaluated the cross-protection afforded by vaccines thatad a poor serological match to heterologous challenge viruses. Thetudy employed vaccines with a very high homologous potencymostly >30 PD50) and found that the heterologous potency wasndeed much less. However, unlike the present study of type O vac-ine, the extent of cross-protection could be predicted from theerological cross-reactivity in that the PA50 was similar for bothomologous and heterologous viruses.

Comparison of the deduced amino acid sequences of the P1 pro-eins of the homologous vaccine, challenge and VN test viruses usedn the present study did not show differences that could accountor the disparity between serologically predicted and actual pro-ection. Aggarwal and Barnett [27] have described five neutralizingntigenic sites of FMDV with critical amino acid residues that formhe neutralizing epitopes. The antigenic site 1 formed by the criti-al residues 144, 148 and 154 and 208 contributed by the �G–�Hoop and carboxy terminus of VP1 did not show any changes. Site

formed in part by residues 43 and 44 of the �B–�C loop of VP1howed no mutations. The fifth site at position 149 of VP1 formedy interaction of the VP1 loop region with other surface amino acidsid not show any mutations. Amino acids at positions 31, 70–73,5 and 77 of VP2 contributing to site 2 remained unchanged. The

ritical residue at 58 of VP3 forming site 4 was unchanged. Com-arison of the deduced amino acid sequence of the capsid codingegion of O1 Manisa and O1 Campos challenge viruses revealed thathe antigenic site 140–153 in the 1D region was different and alsonumber of differences that are not located at known sites of anti-

0.723 0.873 to 3.706 3.17 0.0021.091 −5.684 to −1.405 −3.25 0.0010.620 0.754 to 3.187 3.18 0.001

genic significance, except for at position 159 in site 2 of VP2 wherethe O1 Campos sequence has the soluble amino acid serine whilethat of O1 Manisa has the hydrophobic amino acid proline [28].

Another possible explanation for the discordance betweenin vitro and in vivo findings could be that the O1 Campos challengevirus consists of a mixture of viruses, including a minor compo-nent that was not readily neutralized by O1 Manisa antiserum andthat this virus was selected for during the challenge test. To inves-tigate this possibility, a virus was obtained and characterized froma secondary lesion of a calf that had suffered generalised FMD afterO1 Campos challenge, despite a strong post-vaccination antibodyresponse. Examination of the deduced P1 amino acid sequence ofthis virus revealed that it was homologous to the challenge strainand was still readily neutralized by the cattle sera obtained fromO1 Manisa vaccinated animals of experiment 2 in vitro.

Another possibility is that the O1 Campos challenge was moresevere than the O1 Manisa challenge due to having been givenat a higher dose or as a result of some other inherent virulenttrait. The cattle ID50 titres of the two challenge viruses were thesame, although in terms of TCID50, the Campos titre was one loghigher than the Manisa titre. In the two control groups of unvac-cinated cattle, there is no clear evidence of more extensive virusreplication or virulence in the Campos challenged animals thanin the Manisa challenged animals. Explanations for the mismatchbetween protection predicted by serology and established in vivocould be that the challenge with O1 Campos was more severe thanthe O1 Manisa challenge or that other immune responses to vacci-nation need to be measured to establish an in vitro correlation withcross-protection.

Vaccination of 78 cattle calves resulted in mainly good to mod-erate antibody responses at the time of challenge, although some

animals #761 and #973 in Groups 2 and 3 respectively, #272 and#273 in Group 4 and #337 and #340 in Group 6 in experiment 1and #4092 in Group 5 in experiment 2 did not show any evidenceof sero-conversion. There was a significant difference in the serumneutralizing antibody to O1 Campos and O1 Manisa between exper-

1 / Vacc

it

l3besSpcaaThvactstoass

A

irbTKmh

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[28] Srinivasan VA, Nagendrakumar SB, Madhanmohan M, Maroudam V, Santhaku-mar P, Parida S, et al. Preliminary results to evaluate cross-protection between

912 S.B. Nagendrakumar et al.

ment 1 (P < 0.05) and experiment 2 (P < 0.01) and this could be dueo the difference in the batches of O1 Manisa vaccine.

The antigen payloads in different vaccines in homologous chal-enge showed that the threshold for maximal protection was at.75 �g and decreased thereafter. All the animals vaccinated withoth 15 and 3.75 �g payloads were protected and two animals inxperiment 1 and four in experiment 2 vaccinated with 0.94 �ghowed clinical disease characterized by pyrexia and feet lesions.imilarly, substantially less viral RNA was detected in probang sam-les at 5 dpc from the groups that received 15 and 3.75 �g payloadsompared to those receiving 0.94 �g or no vaccine. In contrast, thenimals vaccinated with 15 and 3.75 �g payloads showed pyrexiand feet lesions when challenged with heterologous O1 Campos.wo animals that were protected in Group 5 of experiment 1 hadigh serum antibody titres for both O1 Manisa and O1 Camposiruses in VNT. Six out of eight animals in experiment 1 and all thenimals in experiment 2 vaccinated with 60 �g of O1 Manisa andhallenged with O1 Campos were protected upon challenge and thewo that were not protected showed pyrexia and feet lesions. Thisuggests that for type O vaccines, there may be benefit in increasinghe antigen payload beyond the threshold for maximum homol-gous strain protection, so as to improve protection in the fieldgainst other strains. All the unvaccinated and challenged animalshowed generalization with feet lesions and pyrexia. These animalshowed a rise in antibody titre from 10 dpc and thereafter.

cknowledgements

Thanks are due to Nick Knowles for the partial VP1 sequenc-ng work and to Pip Hamblin and Bob Statham for serology andvalue work at Pirbright. Work at IAH-Pirbright was supportedy FP6 grant SSPE-CT-2003-503603 and FP7 grant 2009-226556.hanks are also due to M. Manoharan, R. Kumar, J. Anilkumar and. Manikumar for their help in carrying out the animal experi-ents. We thank the anonymous reviewers, whose suggestions

ave helped us to improve the manuscript.

eferences

[1] Brooksby JB. Portraits of viruses: foot-and-mouth disease virus. Intervirology1982;18:1–23.

[2] Cartwright B, Chapman WG, Sharpe RT. Stimulation of heterotypic antigensof foot-and-mouth disease virus antibodies in vaccinated cattle. Res Vet Sci1982;32:338–42.

[3] Rweyemamu MM, Hingley PJ. Foot-and-mouth disease virus strain differenti-ation: analysis of the serological data. J Biol Stand 1984;12:225–9.

[4] Alonso A, Gomes MPD, Ramalho AK, Allende R, Barahona H, Sondahl M, et al.Characterization of foot-and-mouth disease virus by monoclonal antibodies.Viral Immunol 1993;6:219–28.

[5] Kitching RP, Rendle R, Ferris NP. Rapid correlation between field isolates andvaccine strains of foot-and-mouth disease virus. Vaccine 1988;6:403–8.

[6] Paton DJ, Valarcher JF, Bergmann I, Matlho OG, Zakharov VM, Palma EL, et al.Selection of foot-and-mouth disease vaccine strains – a review. Rev Sci TechOIE 2005;24:981–93.

[7] Dunn CS, Samuel AR, Pullen LA, Anderson J. The biological relevance of virusneutralisation sites for virulence and vaccine protection in the guinea pig modelof foot-and-mouth disease. Virology 1998;247:51–61.

ine 29 (2011) 1906–1912

[8] McCullough KC, De Simone F, Brocchi E, Capucci L, Crowther JR, KihmU. Protective immune response against foot-and-mouth disease. J Virol1992;66:1835–40.

[9] Aggarwal N, Zhang Z, Cox S, Statham R, Alexandersen S, Kitching RP, et al. Exper-imental studies with foot-and-mouth disease virus, strain O, responsible for the2001 epidemic in the United Kingdom. Vaccine 2002;20:2508–15.

10] Barteling SJ, Swam H. The potent aqueous and double oil emulsion foot-and-mouth disease type O1 vaccines from European Vaccine Banks probably protectagainst all other O1 strains. Report of the European Commission for the Controlof Foot-and-mouth disease. In: Session of the Research Group of the StandingTechnical Committee, Appendix 13. 2006. p. 90–4.

11] Mattion N, Konig G, Seki C, Smitsaart E, Maradei E, Robiolo B, et al. Reintroduc-tion of foot-and-mouth disease in Argentina: characterisation of the isolatesand development of tools for the control and eradication of the disease. Vaccine2004;22:4149–62.

12] Brehm KE, Kumar N, Thulke H-H, Haas B. High potency vaccines induce protec-tion against heterologous challenge with foot and mouth disease virus. Vaccine2008;26:1681–7.

13] Samuel AR, Knowles NJ. Serological and biochemical study of an isolate offoot-and-mouth disease virus type A from Bavaria in 1984. Rev Sci Tech OIE1988;7:365–73.

14] Doel TR, Mowat GN. An international collaborative study on foot and mouthdisease virus assay methods 2. Quantification of 146S particles. J Biol Stand1985;13:335–44.

15] Rweyemamu MM, Hingley PJ. Food and mouth disease virus strain differentia-tion: analysis of the serological data. J Biol Stand 1984;12:191–4.

16] Snowdon WA. Growth of foot-and-mouth disease virus in monolayer culturesof calf thyroid cells. Nature 1966;210:1079–80.

17] Ferris NP, Dawson M. Routine application of enzyme-linked immunosorbentassay in comparison with complement fixation for the diagnosis of foot-and-mouth and swine vesicular disease. Vet Microbiol 1988;16:201–9.

18] Reid SM, Ferris NP, Hutchings GH, Zhang Z, Belsham GJ, Alexandersen S. Detec-tion of all seven serotypes of foot-and-mouth disease virus by real-time,fluorogenic reverse transcription polymerase chain reaction assay. J Virol Meth-ods 2002;105:67–80.

19] Reid SM, Grierson SS, Ferris NP, Hutchings GH, Alexandersen S. Evaluation ofautomated RT-PCR to accelerate the laboratory diagnosis of foot-and-mouthdisease virus. J Virol Methods 2003;107:129–39.

20] Quan M, Murphy CM, Zhang Z, Alexandersen S. Determinants of earlyfoot-and-mouth disease virus dynamics in pigs. J Comp Pathol 2004;131:294–307.

21] Golding SM, Hedger RS, Talbot P. Radial immuno-diffusion and serum neutral-isation techniques for the assay of antibodies to swine vesicular disease. ResVet Sci 1976;20:142–7.

22] Kärber G. Beitrag zur kollektiven Behandlung pharmakologischer Reihenver-suche. Arch Exp Pathol Pharmakol 1931;162:480–7.

23] Wickham H. Ggplot2: elegant graphics for data analysis (use R). 2nd printinged. New York: Springer-Verlan Inc.; 2009.

24] Gardner MJ, Altman DG, Machin D, Bryant T, Gardner S. Statistics with confi-dence. 2nd revised ed. London: BMJ Publishing; 2000.

25] OIE (Office International des Epizooties/World Organisation for Animal Health).Foot and mouth disease. In: OIE Standards Commission, editor. Manual of stan-dards for diagnostic tests and vaccines. Paris, France: Office International desEpizooties; 2008.

26] Stellmann C, Terre J, Favre H, Brun A, Fontaine J. Comparison of foot and mouthdisease vaccine potency testing on cattle in terms of the nature of the diluent.Arch Virol 1977;54:61–74.

27] Aggarwal N, Barnett PV. Antigenic sites of foot-and-mouth disease virus(FMDV): an analysis of the specificities of anti-FMDV antibodies aftervaccination of naturally susceptible host species. J Gen Virol 2002;83:775–82.

O1 Manisa and O1 Campos in cattle. Appendix 33. International control offoot-and-mouth disease: tools, trends and perspectives. EUFMD Report 2006.Appendix 30. In: Session of the Research Group of the Standing Technical Com-mittee of EUFMD Paphos. 2006.