Institut für Immunologie, Bundesforschungsanstalt für Viruskrankheiten der Tiere, Paul-Ehrlichstr. 28, D Tübingen, Germany

Transcription

1 Journal of General Virology (2003), 84, DOI /vir Induction of an antigen-specific immune response and partial protection of cattle against challenge infection with foot-and-mouth disease virus (FMDV) after lipopeptide vaccination with FMDV-specific B-cell epitopes Bettina-Judith Höhlich, 1 3 Karl-Heinz Wiesmüller, 2 Bernd Haas, 1 Wilhelm Gerner, 1 Roberto Correa, 3 Hans-Robert Hehnen, 3 Tobias Schlapp, 3 Eberhard Pfaff 1 and Armin Saalmüller 1 Correspondence Armin Saalmüller armin.saalmueller@tue.bfav.de 1 Institut für Immunologie, Bundesforschungsanstalt für Viruskrankheiten der Tiere, Paul-Ehrlichstr. 28, D Tübingen, Germany 2 EMC microcollections GmbH, Sindelfinger Str. 3, D Tübingen, Germany 3 Bayer Animal Health, Building 6210, D Leverkusen, Germany Received 22 May 2003 Accepted 19 August 2003 To evaluate the potential of chemically synthesized lipopeptides for vaccination against foot-and-mouth disease (FMD), seven lipopeptides containing the immunostimulating principle of bacterial lipoproteins and linear B-cell epitopes of FMDV strain O 1 Kaufbeuren (O 1 K) were used to immunize cattle (n=7). Animals were vaccinated once and 21 days after immunization animals were infected with the homologous virus. Four animals were protected. After vaccination, as well as after challenge infection, B- and T-cell responses were examined. Sera were tested for virus- and peptide-specific antibodies and showed after vaccination only a weak antibody response. After challenge infection, an increase in antibody titre was obvious but there was no correlation between antibody titre and protection. The reactivity of the cellular immune system was detected by analyses of PBMCs for virus- and peptide-specific T-lymphocytes. With regard to the virus-specific T-lymphocytes, a heterogeneous reactivity could be shown. No correlation between virus-specific T-cell proliferation and protection was found. Obvious was the fact that all protected animals showed after vaccination a strong T-cell response against at least one of the peptides used for immunization. These results suggest a correlation between the onset of an antigen-specific T-cell reaction and protection. INTRODUCTION Foot-and-mouth disease (FMD), a devastating disease of livestock, is caused by FMD virus (FMDV), a member of the genus Aphthovirus within the family Picornaviridae. Although animals will generally recover from the acute clinical phase of the disease, usually there is a permanent loss of performance, due mostly to chronic lameness, permanent drop in milk yield and poor weight gain. FMD is the most contagious virus disease of animals (Pereira, 1981) and is endemic in many regions of the world, including much of Africa, Asia and some countries in South America. In FMD-endemic regions, the major cost of the disease is associated with reduced livestock productivity and reduced access to international markets for 3Present address: Miltenyi Biotec, Friedrich-Ebert-Str. 68, D Bergisch Gladbach, Germany. livestock and livestock products. In many situations, regular vaccination is an essential part of the disease control strategy. Regions free of FMD include Europe, Australia, Japan and North and Central America. Here, control is based upon prevention of introduction of the virus through rigorous importation regulation and quarantine. In case of a disease introduction into these areas, emergency vaccination may have to support movement controls and stamping out in order to eradicate the disease. Current FMD vaccines are produced by infecting cultured cells with virulent FMDV, followed by inactivation and purification of the newly synthesized virus particles. Whereas the resulting conventional vaccines have been applied successfully for decades, they have a number of disadvantages. Handling FMDV is inherently hazardous and necessitates high security facilities (Barteling & Vreeswijk, 1991). To ensure reliable inactivation, FMD vaccine plants G 2003 SGM Printed in Great Britain 3315

2 B.-J. Höhlich and others need a special design and complex operating procedures that have to be strictly monitored. A conventional FMD vaccine is a biological product that is very difficult to standardize. To ensure the required potency, animal experiments are necessary, not only for vaccine registration but also for batch control (Barteling & Vreeswijk, 1991). In addition, vaccines will quickly lose potency without a consistent cold chain from the producer to the veterinarian in the field. Furthermore, protection induced by conventional vaccines is limited mostly to one or a few topotypes within one serotype (Pereira, 1981). New strains are constantly evolving in endemically infected countries, making it difficult for vaccine producers to keep pace. Protection is short-lived, often making frequent revaccination indispensable for an effective disease control (Barteling & Vreeswijk, 1991). Whereas conventional vaccines can prevent clinical signs and further spread of disease in vaccinated populations, they are not able to protect virus-exposed animals from becoming persistently infected virus carriers. To overcome these problems, a lot of work was performed on alternatives to conventional vaccines. Synthetic peptides were early found to be promising vaccine candidates (Pfaff et al., 1982; DiMarchi et al., 1986; Doel et al., 1990). Major antigenic regions of the virus were found on the structural protein 1D (Pfaff et al., 1982; Strohmaier et al., 1982; Acharya et al., 1989; Bittle et al., 1982). The antigenic site comprised of aa of 1D (containing the receptorbinding site of the virus) is usually referred to as site A, whereas site C contains aa of 1D. In initial experiments, peptide vaccines consisting of site A or of constructs combining sites A and C were used. Whereas these vaccines induced considerable titres of neutralizing antibodies in guinea pigs, cattle and swine and sometimes also good protection in vivo (Pfaff et al., 1982; Bittle et al., 1982; DiMarchi et al., 1986; Morgan & Moore, 1990), in general, their immunogenicity was lower than that of conventional vaccines. This may be due to the now widely accepted thesis: that the immune response against FMDV is not only B-cell but also T-cell dependent (Domingo et al., 1990; Collen et al., 1991; Brown, 1992; Collen, 1994; Sobrino et al., 1999). Because sites A and C contain mainly B-cell epitopes, in a recent study these immunogenic sites were combined with a further 1D peptide containing a T-cell epitope (Taboga et al., 1997). This resulted in a minor increase in protection rates. The aim of this study was to evaluate the potential of lipopeptides containing FMDV-specific B-cell epitopes from structural and non-structural (NS) proteins to induce an FMDV-specific immune response in cattle. A special focus was laid on the variance of the immune response of different randomly selected outbred cattle against the peptides used for the immunizations. In addition, the potency of the B-cell epitopes in the induction of a peptide and/or FMDV-specific T-cell reaction was studied. Furthermore, the capacity of the lipopeptides to induce protection against a homologous challenge infection was investigated. METHODS Animals. All cattle used for vaccination experiments were approximately 1 year old and free of antibodies against FMDV, as confirmed by liquid-phase blocking sandwich ELISA (LPBE) (Hamblin et al., 1986). Sera of cattle infected with FMDV strain O 1 Lausanne (O 1 L) were kindly provided by M. Parkhouse, Institute for Animal Health, Pirbright, UK. Sera of cattle infected with FMDV strains A 5 Bernbeuren (A 5 B), Asia 1 Shamir (A 1 S) and SAT 1 Zimbabwe (SAT 1 Z) were obtained from infected animals between 10 and 28 days after challenge infection. Viruses. FMDV strain O 1 Kaufbeuren (O 1 K) was grown on monolayers of BHK cells, clone Tübingen. Virus stocks used for proliferation assays were stored as clarified tissue culture harvest material (3000 g for 20 min) at 270 uc. For LPBE (Hamblin et al., 1986), virus was inactivated with binary ethyleneimine and additionally incubated at 32 uc overnight. Virus used for challenge infections was passaged in cattle and each animal was infected with p.f.u. intralingually. Peptides. Peptides were synthesized by conventional solid-phase methodology (Merrifield & Stewart, 1965) and dissolved in a concentration of 10 mg ml 21 in DMSO. Overlapping peptide amides (14 or 15 aa in length, overlapping each other by 10 aa) were synthesized based on the sequence of FMDV O 1 K(Forsset al., 1984) using Fmoc chemistry (Multiple Peptide Synthesizer SyRo, MultiSyntech). Synthesized peptides were analysed by analytical high-pressure liquid chromatography (System Gold, Beckman Coulter) on a Nucleosil C18 column (Grom) and ion-spray mass spectroscopy (VG Quattro II Triple-Quatrupol Ionspray-MS, Micromass). The purities of crude peptides were, in general, higher than 70 %. For immunization purposes, peptides were synthesized as Pam 3 Cys peptides, as described previously (Wiesmüller et al., 1989). Vaccination and infection of animals. Vaccination was carried out intramuscularly with 2 mg peptide per animal. Lipopeptides were dissolved in PBS and emulsified with an equal volume of Montanide ISA 9. A 1?5 ml volume per dose was administered. Challenge infections were carried out according to the methodologies described by the European Pharmacopoeia for the determination of the potency of conventional vaccines. A total of 10 4 p.f.u. FMDV O 1 K was inoculated intradermally into two sites on the upper surface of the tongue (0?1 ml per site). After 8 days, animals were sedated with Rompun and examined carefully for FMD-specific lesions. Animals that showed lesions at sites other than the tongue (secondary lesions) at 8 days post-infection were considered nonprotected. To control the challenge infection, two control animals were included in the experiment. Both control animals developed lesions: one animal showed FMDV-caused lesions on all four extremities, the other one only on two extremities. LPBE. LPBE was performed as described (Hamblin et al., 1986), with slight modifications. Briefly, ELISA plates (Nunc) were coated overnight with rabbit anti-fmdv O 1 K serum. Twofold dilution series of each test serum were mixed with equal volumes of antigen (cell culture supernatant) of a predetermined dilution and kept at 4 uc overnight in a dummy plate. After washing, each well of the coated ELISA plates received 50 ml antigen/serum mixture. Control wells received antigen/buffer mixture (antigen control) or antigen/ standard serum mixture, respectively. After incubation for 1 h at 37 uc, plates were washed and incubated with guinea pig anti- FMDV O 1 K serum. After incubation for 1 h at 37 uc, plates were washed again and incubated for a further 1 h at 37 uc with goat anti-guinea pig gamma-globulin-peroxidase conjugate (Biozol). After the final washing step, the chromogen orthophenylenediamine (OPD) (Sigma) with H 2 O 2 was added. The reaction was stopped 3316 Journal of General Virology 84

3 FMDV-specific B-cell epitopes after 15 min with H 2 SO 4 and plates were read in an automatic microplate reader at 492 nm. Antibody titres were expressed as the dilution at which the test sera showed greater than 50 % inhibition of the absorbance value recorded for the antigen control wells. Titres below 1 : 40 (1?6 log 10 ) were considered positive. Peptide ELISA. Plates (Immunoplate Maxisorb, Nunc) were coated with 10 mg peptide ml 21 in 100 ml H 2 O per well and air-dried at 37 uc. Peptide-coated plates were blocked for 2 h at 37 uc with 3 % BSA/PBS and then incubated with sera diluted in 3 % BSA/PBS at a 1 : 100 dilution for 1 h at 37 uc. Specific antibodies were detected with horseradish peroxidase-conjugated goat anti-bovine immunoglobulin antibodies (Dianova). Colour development was obtained after addition of the substrate (OPD with 0?2 ml H 2 O 2 ml 21 ). The reaction was stopped with 2 M H 2 SO 4. All plates were read in an automatic microplate reader (Titertek). Results are expressed as indices (average absorbance value of duplicate wells of the sample/average absorbance value of the duplicate wells of the negative controls). As negative controls, three FMDV O 1 K peptides that did not contain linear B-cell epitopes were used. The indices of all animals were below 2 before the first vaccination. Lymphocyte proliferation assay. Bovine PBMCs were isolated by Histopaque (Histopaque 1083, Sigma) centrifugation of heparinized blood. Cultivation of PBMCs ( ml 21 ) was performed in MEM-a (Gibco-BRL) supplemented with 10 % FCS, 2 mm L-glutamine, M 2-mercaptoethanol, 10 mm HEPES buffer, 100 IU penicillin ml 21 and 0?1 mg streptomycin sulphate ml 21. For the proliferation assay, PBMCs were incubated for 5 days at 37 uc with 5 % CO 2 in 200 ml microcultures in round-bottomed microtitre plates (Greiner). PBMCs from vaccinated animals were restimulated with either FMDV O 1 K (log 2 titration, starting with p.f.u. per well) or synthetic peptides (25 and 2?5 mg ml 21 ). As negative controls, cell culture supernatants from mock-infected cells or irrelevant 15-mer peptides derived from hepatitis C virus dissolved in DMSO (25 and 2?5 ml ml 21 ) were used. To quantify the proliferative response, 1 mci [ 3 H]thymidine (Amersham) per well was added for an additional h. After the microplates were freeze-thawed, the cells were harvested using a multiharvester system (Wallac) and [ 3 H]thymidine uptake was measured in a b-counter (Wallac). All results are expressed as stimulation indices (SI), which can be defined as geometric mean c.p.m. of the triplicate microcultures of the sample/geometric mean c.p.m. of the triplicate microcultures of the respective negative controls. The SI determined before vaccination was below 2 for all animals. RESULTS Identification of FMDV-specific linear B-cell epitopes FMDV-specific linear B-cell epitopes were identified by screening sera of FMDV O 1 K-infected cattle for peptidespecific antibodies in a peptide ELISA, as mentioned above. Fifteen of 442 tetra- or pentadeca peptides, which overlapped in 10 aa and spanned the whole ORF of FMDV O 1 K, could be identified as linear B-cell epitopes. As demonstrated in Fig. 1, these epitopes were distributed over five NS proteins (2B, 2C, 3A, 3B and 3D) and were also located in the RGD region of structural protein 1D (Fig. 1). This region was characterized as the receptor-binding site of FMDV and is known as the major immunogenic site of the virus. The sequences of the peptides identified as FMDVspecific linear B-cell epitopes are presented in Fig. 2. Analysis of the reactivity of the linear B-cell epitopes derived from FMDV O 1 K with sera from cattle infected with other FMDV serotypes Because of the fact that most of the peptides identified as linear B-cell epitopes for strain O 1 K derived from highly conserved regions of FMDV, their reactivities with sera from 21 animals infected with other FMDV sub- and serotypes were studied (Fig. 3). In these experiments, in addition to sera of FMDV O 1 K-infected cattle, sera of cattle infected with FMDV strains O 1 L, A 5 B, A 1 S and SAT 1 Z were examined also. Fig. 1. Localization of FMDV O 1 K-specific linear B-cell epitopes for cattle. Schematic map of the FMDV ORF. Indicated are the two NS protein regions encoding the NS proteins L/L9, 2A 2C and 3A 3D. The structural protein region encoding the capsid proteins 1A 1D is also shown. Within the ORF the localization of the peptides, identified as linear B-cell epitopes (bars), is shown. Epitopes are numbered consecutively from the N-terminal end of the sequence. Synthetic peptides used for the identification of the linear B-cell epitopes were 14 to 15 aa long with an overlap of 10 aa and represent the amino acid sequence of FMDV O 1 K

4 B.-J. Höhlich and others Fig. 2. Sequence of FMDV O 1 K-specific linear B-cell epitopes. The amino acid sequences of peptides that contain linear B-cell epitopes for FMDV O 1 K are shown. Additionally, the viral protein from which the sequence derived is indicated. The numbers of the peptides refer to Fig. 1. For immunization experiments, only the peptides indicated as being positive (+) were used. As expected, all FMDV strains were able to induce antibodies against several linear B-cell epitopes identified before. All sera tested so far shared common linear B-cell epitopes. But the B-cell response against epitopes derived from the regions of the structural protein 1D, and the NS proteins 2B, 2C, 3A and 3D (peptides 1 9 and 15), showed within the group of animals great virus strain-independent animal-to-animal variation. This could be demonstrated for the group of FMDV O 1 K-infected animals (Fig. 3, lines 1 4) as well as for all other groups. Interestingly, most epitopes derived from the 3B region (peptides 10 14) were recognized more consistently within the groups. In this context, it should be mentioned that protein 3B consists of three nonidentical tandem repeats of the viral genome-associated protein (VPg) and the peptides derived from this protein share parts of their sequence. Peptides 10 and 11 belong to the first tandem repeat and show an overlap in 9 aa (PLERQKPLK), peptides 12 and 13 are based on the equivalent structures of the second repeat and share 9 aa also (PMERQKPLK). Interestingly, peptide 13, which showed no difference between FMDV O 1 K and O 1 L in its sequence, was recognized less by the sera from the FMDV O 1 L- infected animals. Peptide 14, recognized by a more limited number of sera, is part of the third repeat, being equivalent to peptides 11 and 13. This third repeat in the 3B region is the most heterogeneous part of the protein. The most remarkable fact revealed by this comparison of the immune response against FMDV-specific linear B-cell epitopes is that each serum derived from an infected animal that was tested elicited antibodies against both, or, in the case of the animals #38 and #296, at least against one, of the overlapping peptides 10 and 11. The peptides belonging to the second and third VPg repeat were recognized less efficiently. Immunization of cattle with lipopeptides and determination of their antibody responses To determine the ability of the identified linear B-cell epitopes to induce an in vivo immune response and to protect immunized animals from challenge infection, peptides identified as linear B-cell epitopes (Fig. 2) were selected, synthesized and N-terminally elongated by the immunostimulating lipoamino acid Pam 3 Cys. To reduce the number of peptides for the immunization experiments, only seven peptides were selected out of the 15 identified as linear B-cell epitopes (Figs 1 and 2), namely peptides 1, 3, 4, 6, 7, 11 and 15. Peptide 1 derived from the 1D structural protein contained the RGD region responsible for the attachment of the virus to the cellular receptor. The epitope corresponding to peptide 3 is located in the C-terminal part of NS protein 2B. Peptides 4 and 6 describe two B-cell epitopes located in the N terminus of NS protein 2C. Peptide 7 was derived from protein 3A and peptide 11 from the first tandem repeat of protein 3B. Peptide 15 was part of the N terminus of protein 3D. The sequences of all these peptides used for the immunization experiments are summarized in Fig. 2. As mentioned above for the vaccination experiments, these peptides were synthesized as lipopeptides and applied in an oil emulsion. Seven cattle were immunized with a dose of 2 mg for each peptide per animal. Blood samples were taken from each animal 21 days 3318 Journal of General Virology 84

5 FMDV-specific B-cell epitopes Fig. 3. Reactivity of linear B-cell epitopes from FMDV O 1 K with sera from cattle infected with different FMDV sub- and serotypes. The existence of antibodies against peptides containing linear B-cell epitopes in sera of cattle infected with FMDV strains O 1 K, O 1 L, A 1 S, A 5 B and SAT 1 Z is shown. Sera were examined in a peptide ELISA (see Methods). Serum samples were deemed positive for a peptide when the sample reached an SI of at least 2. later and the induction of specific antibodies against FMDV and against the synthetic peptides used for the immunization was determined. Animals were challenged with FMDV O 1 K, from which the sequence of the synthetic peptides had been derived. In this experiment, four of seven animals were protected (Fig. 4b). To characterize the effects of lipopeptide vaccination and infection on the immune response of the respective animals, the humoral immune response was investigated 21 days after vaccination and days after challenge infection (Fig. 4). The specific antibody reactivity of non-protected (Fig. 4a) and protected (Fig. 4b) animals is compared in Fig. 4. After vaccination, significant antibody titres against whole virus could be detected in neither the sera of unprotected (Fig. 4a) nor the sera of protected animals (Fig. 4b). After the challenge infection, all animals showed a clear antibody titre against FMDV in the LPBE. Interesting is the fact that the antibody titres of the protected animals (Fig. 4b) were, on average, lower compared to the titres of the non-protected animals (Fig. 4a). Investigating the sera derived after the first immunization prior to the challenge infection for antibodies against peptides (Fig. 4, closed bars), the sera of most animals showed a very low reaction in the peptide ELISA and only against a few of the peptides that were included in the vaccine (A 492 >2). The serum of animal #69064 showed a response to peptide 15 and the serum of animal #99219 showed a response against peptides 1 and 7. However, there were also animals that did not elicit a detectable amount of antibodies against any of the peptides, for example, the protected animal # In short, the humoral immune response after vaccination revealed no conclusive pattern. About 3 weeks after challenge (Fig. 4, open bars), the sera of most animals showed an increased antibody level not only against the peptides they had already responded to after the vaccination but also against peptides they did not show any response to before the vaccination. Protected animal #15879 had developed antibodies against peptides 7 and 15 after vaccination and additionally against peptides 1, 3, 6 and

6 B.-J. Höhlich and others Fig. 4. B-cell response of cattle after vaccination with synthetic lipopeptides and after challenge infection. Sera of seven cattle vaccinated with a single dose of synthetic lipopeptides (see Fig. 2) were obtained 21 days after vaccination and 17 (animals #43912 and #69064), 21 (animals #15879, #38715 and #99219) or 22 (animals #66671 and #71775) days after challenge infection. Sera were tested for antibodies against whole virus and against the peptides that were contained in the vaccine. The titres of antibodies against whole virus are presented on the right of the respective individual diagrams. The upper rows marked by ( a ) indicate the antibody titre before and the numbers in the second rows marked by ( b ) indicate the antibody titre after infection. The results obtained in the peptide ELISA (closed bars, after vaccination; open bars, after challenge) are presented as diagrams. The horizontal lines represent A 492 =1. All analyses are based on three independent experiments with two replicates. SD values in all experiments were less than 10 %. The animals were divided into two groups: non-protected (a) and protected (b) animals. after challenge infection. An exception was animal #43912, which showed, in contrast to all other animals, almost no detectable antibody response against any peptide in the peptide ELISA. This animal had also a very low antibody titre in the LPBE after the first immunization, but after the challenge infection the antibody titre was quite high (titre>1 : 2400). A comparison of the antibody responses of the four protected and the three non-protected animals showed no clear difference, neither with regard to the reactivity of the sera to FMDV in the LPBE nor in the peptide ELISA. A minor point might be that the protected animals showed on average a higher antibody level against the peptides after challenge than the non-protected animals, but animal #71775 was not protected in spite of a good antibody response to several peptides. A summary of the data about the humoral immune response of lipopeptide-vaccinated cattle against FMDV and against synthetic peptides as described in detail above appears to be difficult and it seems to be impossible to make a clear correlation between antibody response and protection against a challenge infection. Within the group of protected and also within the group of non-protected animals, animals with high and low antibody levels against peptides as well as FMDV could be found. In vitro analysis of the cellular immune response Because no clear correlation between the humoral immune response and protection could be recognized in the experiments described above, further investigations were directed towards the second branch of the immune system the 3320 Journal of General Virology 84

7 FMDV-specific B-cell epitopes Fig. 5. T-cell response of cattle after vaccination with synthetic lipopeptides and after challenge infection. The T-cell response of the animals introduced in Fig. 4 was determined after vaccination and after challenge infection (same time points as for B-cell response) in a proliferation assay (see Methods; closed bars, after vaccination; open bars, after challenge). SI values shown were evaluated as the average SI of three independent assays. SD values in all assays were less than 15 %. SI values were calculated by dividing c.p.m. of the FMDV peptide- or virus-stimulated group by the c.p.m. of the control groups stimulated with either irrelevant peptides (hepatitis C virus) or supernatant from FMDV mock-infected cells. The c.p.m. of the control groups were in all experiments less than A clear reactivity was defined when the SI was more than 5 (horizontal lines). cellular immune response mediated by antigen-specific T-lymphocytes. Therefore, PBMCs of all vaccinated animals were examined in lymphoproliferation assays for their virus- and peptide-specific T-cell response after lipopeptide vaccination and after challenge infection. Most of the cattle showed a virus-specific T-cell response after immunization (Fig. 5, closed bars). Both groups, protected and non-protected animals, contained high (e.g. animals #69064 and #15879) as well as weak (e.g. animals #43912 and #99219) responding animals. The response of the bovine T-lymphocytes against the B-cell epitopes used for immunization was investigated also. For this, PBMCs of vaccinated animals as well as PBMCs derived from the same animals after challenge infection were restimulated in vitro with the synthetic peptides used for immunization. With regard to the peptide-specific T-cell response after vaccination, a big variability of response could be detected in both protected and non-protected animals (Fig. 5). All peptides used for the immunizations were able to elicit a significant T-cell response (SI>5) in at least one of the animals, but the pattern of the response was very heterogeneous and differed from animal to animal. In the non-protected group (Fig. 5a), peptide 6 (NP protein 2C) was able to induce a clear in vitro T-cell response in animal # In the protected group (Fig. 5b), PBMCs from animal #15879 showed a high response against peptide 3 from the 2B region. The same was true for animal # In contrast to these two animals, animal #99219 showed a high response against peptide 7 from the 3A region, whereas #66671 showed reactions with peptides 1 (1D), 6 (2C) and 15 (3D). No positive correlation could be found between the peptidespecific reaction pattern of the respective T-cells and protection. But concerning the intensity of the T-cell reactions a correlation with protection can be suggested. All protected animals showed a very high proliferative response (SI>15) against at least one of the peptides. Animals #15879 and #38715, for example, showed a very high response to peptide 3, whereas animals #99219 and #66671 reacted with a high proliferative response towards peptides 7 and 15, respectively. None of the non-protected animals showed such a high T-cell activity against any of the peptides before challenge infection (Fig. 5a)

8 B.-J. Höhlich and others capsid-forming structural proteins, the NS proteins show a considerably higher conservation in their amino acid sequence. The most consistent results in the tests for the reactivity of the antibodies could be found for peptides 10 and 11, derived from the 3B protein. All FMDV-infected cattle tested so far showed an antibody response against at least one of these two overlapping peptides. The other peptides showed a higher inconsistency in their reactivity but were also recognized by sera of cattle infected with FMDV serotypes others than O 1 K, which served as basis for the peptide synthesis. This may lead to the conclusion that mainly individual differences between animals are responsible for the various recognition patterns of linear B-cell epitopes and not only the FMDV serotype with which they were infected. Fig. 6. Summary of the peptide-specific T- and B-cell response of cattle induced by lipopeptides. The peptide-specific T- and B-cell response of the cattle introduced in Figs 4 and 5, before challenge infection, in simplified terms. The B-cell response was regarded as negative ( ) when the serum contained not enough antibodies against any peptide, contained in the vaccine, to reach an SI of 3 (see Methods). The T-cell response of an animal was then regarded as negative when the PBMCs did not show an SI over 5 to any peptide, contained in the vaccine, in the proliferation assay. Could a SI over 5 or over 15 be detected for at least one peptide the response was determined as weak (+) or good (++), respectively. In addition to the B- and T-cell response, it is indicated if the animal was protected (+) or non-protected ( ). A strategy to overcome confusion about immune response and protection might be a combinatorial view on B- and T-cell responses of the protected and non-protected animals. Maybe a summary of the B- and T-cell responses, as presented in Fig. 6, can give an explanation for a correlation between the FMDV-specific immune response and protection. Looking at Fig. 6, in which the antigen-specific B- and T-cell responses of all animals are summarized, it becomes more and more obvious that all animals showing a high T-cell response after the peptide immunization were protected, independent of their B-cell response. DISCUSSION For this study, the whole ORF of FMDV O 1 K was examined for linear B-cell epitopes. Sections of 14 or 15 aa (overlapping in 10 aa) were analysed and 15 linear B-cell epitopes located in the structural as well as in the NS protein region of FMDV could be detected. For several of the peptides a broad reactivity with sera from cattle infected with different other sero- and subtypes of FMDV could be shown. This can be explained easily by the fact that most of the peptides (2 15) originate from NS proteins. Compared to the viral These results indicate that if it should be possible to formulate a protective vaccine containing these B-cell epitopes, a good chance would exist that this vaccine would show protection against several or most of the FMDV sero- and subtypes. The idea to use peptides for the development of a FMDV vaccine already existed in the eighties of the last century. Different B- and T-cell epitopes were used for protection experiments with cattle (DiMarchi et al., 1986; Morgan & Moore, 1990; Taboga et al., 1997; Volpina et al., 1999). The results varied strongly. The ratio of protected animals in some experiments was quite high in spite of the fact that in most cases no more than two different peptides were used for vaccination. But there were already hints that the use of only few B- and T-cell epitopes supports the selection of escape variants (Taboga et al., 1997). As in all of these early experiments, all peptides were derived from the structural proteins of FMDV, a heterotypic protection against different FMDV serotypes appeared to be highly unlikely. In contrast to these experiments, the study presented here used a vaccine with seven different epitopes (six of seven) derived mainly from the NS protein region. Regarding the peptide-specific B-cell response, vaccinated animals showed a reaction pattern against single peptides that was highly variable. Such variations in antigen recognition are already known for FMDV immunizations (McCullough et al., 1992). Even if a B-cell response is not MHC-restricted, quite often animal-to-animal variations are observed for the same antigen (Riley, 1996). In contrast to peptide-specific antibodies that could be detected in the sera of most vaccinated cattle, only in few animals could antibodies against FMDV be determined with the ELISA technique used commonly for FMDV diagnosis (Hamblin et al., 1986). When looking at the cellular immune response, most cattle showed a good reaction against single peptides after vaccination. Astonishing was the extremely good proliferative response that could be shown for some of the protected 3322 Journal of General Virology 84

9 FMDV-specific B-cell epitopes animals (SI>15). As mentioned above, a correlation with the antibody titre could not be presented in these animals. But despite the absence of detectable antibody titres against FMDV, these animals were protected from challenge infection. This is most probably due to an effective T-cell reaction that was able to support the formation of very effective neutralizing antibodies. On the other hand, this protection by T-lymphocytes might be explained by an effective cellular immune response, including the activity of T-helper cells as well as cytolytic T-lymphocytes. Further studies might elucidate these speculations. It was also interesting to note with regards to the T-cell response against the synthetic peptides that five of the seven peptides used for the immunization experiments seemed to contain T-cell epitopes responsible for the in vitro restimulation of the respective T-lymphocytes. Lipopeptides are synthetic analogues of membrane lipoproteins of bacteria and activate cells for cytokine release by their interaction with Toll-like receptors, which indicates a molecular link between host defence mechanisms and microbial products. (Aliprantis et al., 1999). Synthetic analogues of the N-terminal part of these lipoproteins constitute potent immunoadjuvants in vitro and in vivo (Wiesmüller et al., 1992). One might also speculate that the Pam 3 Cys-coupled peptides were caught by the specific B-cell receptors, the surface immunoglobulin molecules, internalized into the B-cells and presented by B-cell MHC molecules to the respective T-cell receptors of the specific T-lymphocytes. At this stage, it is not clear whether the presentation is due mainly to MHC class II molecules as it might be expected or whether MHC class I molecules are involved also. Nothing is known to date whether the Pam 3 Cys-anchor molecule can be involved in a kind of cross-presentation. These questions will be the content of further studies. Another peculiarity without clear explanation was the reduced proliferation against peptides after challenge infection. The reason for this decrease of the proliferative capacity is most probably due to the fact that the ability of the T-cells to become restimulated was inhibited due to the strong stimulation before. This could be a momentary effect, as described similarly for the mouse (De Mattia et al., 1999), but such a decreased reactivity of antigen-specific T-helper cells after restimulation could also occur after restimulation with the same antigen after consecutive vaccinations. In general, it can be concluded that it seems that animals were protected (four of seven) when a high T-cell activity (SI>15) could be shown for at least one peptide. This information is important for a further optimization of a potential peptide vaccine, and after a detailed analysis of this ex vivo data, it might be possible in future experiments to avoid challenge infections. This would considerably reduce the cost and time needed for further investigations. ACKNOWLEDGEMENTS The authors thank Bayer Leverkusen for funding, Andreas Moss for help with the animal experiments and Gabriele Kuebart and Thorsten Decker for technical help. REFERENCES Acharya, R., Fry, E., Stuart, D., Fox, G., Rowlands, D. J. & Brown, F. (1989). The three-dimensional structure of foot-and-mouth disease virus at 2?9 Å resolution. Nature 337, Aliprantis, A. O., Yang, R. B., Mark, M. R., Suggett, S., Devaux, B., Radolf, J. D., Klimpel, G. R., Godowski, P. & Zychlinsky, A. (1999). Cell activation and apoptosis by bacterial lipoproteins through Tolllike receptor-2. Science 285, Barteling, S. J. & Vreeswijk, J. (1991). Developments in foot-andmouth disease vaccines. Vaccine 9, Bittle, J. L., Houghten, R. A., Alexander, H., Shinnick, T. M., Sutcliffe, J. G., Lerner, R. A., Rowlands, D. J. & Brown, F. (1982). Protection against foot-and-mouth disease by immunization with a chemically synthesized peptide predicted from the viral nucleotide sequence. Nature 298, Brown, F. (1992). New approaches to vaccination against foot-andmouth disease. Vaccine 10, Collen, T. (1994). Foot-and-mouth disease virus (aphthovirus): viral T cell epitopes. In Cell-Mediated Immunity in Ruminants, pp Edited by B. M. I. Goddeeris & I. Morrison. Boca Raton: CRC Press. Collen, T., DiMarchi, R. & Doel, T. R. (1991). A T cell epitope in VP1 of foot-and-mouth disease virus is immunodominant for vaccinated cattle. J Immunol 146, De Mattia, F., Chomez, S., Van Laethem, F., Moulin, V., Urbain, J., Moser, M., Leo, O. & Andris, F. (1999). Antigen-experienced T cells undergo a transient phase of unresponsiveness following optimal stimulation. J Immunol 163, DiMarchi, R., Brooke, G., Gale, C., Cracknell, V., Doel, T. & Mowat, N. (1986). Protection of cattle against foot-and-mouth disease by a synthetic peptide. Science 232, Doel, T. R., Gale, C., Do Amaral, C. M. C. F., Mulcahy, G. & DiMarchi, R. (1990). Heterotypic protection induced by synthetic peptides corresponding to three serotypes of foot-and-mouth disease virus. J Virol 64, Domingo, E., Mateu, M. G., Martínez, M. A., Dopazo, J., Moya, A. & Sobrino, F. (1990). Applied Virology Research, vol. 2, pp Edited by E. Kurstak, R. G. Marusyk & B. Murphy. New York: Plenum. Forss, S., Strebel, K., Beck, E. & Schaller, H. (1984). Nucleotide sequence and genome organization of foot-and-mouth disease virus. Nucleic Acids Res 12, Hamblin, C., Barnett, I. T. & Hedger, R. S. (1986). A new enzymelinked immunosorbent assay (ELISA) for the detection of antibodies against foot-and-mouth disease virus. I. Development and method of ELISA. J Immunol Methods 93, McCullough, K. C., Bruckner, L., Schaffner, R., Fraefel, W., Müller, H. K. & Kihm, U. (1992). Relationship between the anti-fmd virus antibody reaction as measured by different assays, and protection in vivo against challenge infection. Vet Microbiol 30, Merrifield, R. B. & Stewart, J. M. (1965). Automated peptide synthesis. Nature 207, Morgan, D. O. & Moore, D. M. (1990). Protection of cattle and swine against foot-and-mouth disease, using biosynthetic peptide vaccines. Am J Vet Res 51,

10 B.-J. Höhlich and others Pereira, H. G. (1981). Virus Diseases of Food Animals, pp Edited by E. P. Gibbs. London, UK: Academic Press. Pfaff, E., Mussgay, M., Böhm, H. O., Schulz, G. E. & Schaller, H. (1982). Antibodies against a preselected peptide recognize and neutralize foot-and-mouth disease virus. EMBO J 1, Riley, E. M. (1996). The role of MHC- and non-mhc-associated genes in determining the human immune response to malaria antigens. Parasitology 112, Sobrino, F., Blanco, E., Garcia-Briones, M. & Ley, V. (1999). Synthetic peptide vaccines: foot-and-mouth disease virus as a model. Dev Biol Stand 101, Strohmaier, K., Franze, R. & Adam, K. H. (1982). Location and characterization of the antigenic portion of the FMDV immunizing protein. J Gen Virol 59, Taboga, O., Tami, C., Carrello, E. & 13 other authors (1997). A large-scale evaluation of peptide vaccines against foot-and-mouth disease: lack of solid protection in cattle and isolation of escape mutants. J Virol 71, Volpina, O. M., Surovoy, A. Y., Zhmak, M. N., Kuprianova, M. A., Koroev, D. O., Chepurkin, A. V., Toloknov, A. S. & Ivanov, V. T. (1999). A peptide construct containing B-cell and T-cell epitopes from the foot-and-mouth disease viral VP1 protein induces efficient antiviral protection. Vaccine 17, Wiesmüller, K.-H., Jung, G. & Hess, G. (1989). Novel low-molecularweight synthetic vaccine against foot-and-mouth disease containing a potent B-cell and macrophage activator. Vaccine 7, Wiesmüller, K.-H., Bessler, W. G. & Jung, G. (1992). Solid phase peptide synthesis of lipopeptide vaccines eliciting epitope-specific B-, T-helper and T-killer cell response. Int J Pept Protein Res 40, Journal of General Virology 84