FEMS Immunology and Medical Microbiology 10 (1995) 191-198 0 1995 Federation of European Microbiological Societies 0928-8244/95/$09.50 Published by Elsevier

FEMSIM 00474

Generatio:n of rubella virus-neutralising antibodies by vaccination with synthetic peptides

191

Karen Robinson *, Ann Mostratos and Richard K. Grencis

School of Biological Sciences, Universiry of Manchester, Manchester Ml3 9PT, UK

(Received 27 September 1994; revision received 9 November 1994; accepted 10 November 1994)

Abstract: Four short peptides from rubella virus proteins El and E2, predicted to contain B cell epitopes, were used to vaccinate BALB/c mice. Sera from peptide-vaccinated animals reacted with viral antigens in ELISA and three of the four induced virus-neutralising antibody (nAb) responses. Peptide PY4, in contrast to the others, induced IgG2a responses upon vaccination and stimulated spleen cells in vitro produced IFNy in the absence of IL-5. It was reasoned that vaccination with PY4 caused Thl subset activation, the appropriate type of response for anti-viral immunity and hence the efficient neutralising antibody response. Presentation of peptide -for vaccination proved to be as important as the sequence. Similar profiles of IgGl and IgG2a were detected in the sera of mice vaccinated with PY4 in Freund’s complete adjuvant or alum; however nAb responses were not found when alum was used.

Key work Rubella; Synthetic peptide; Vaccination; Neutralising antibody; Adjuvant; Epitope prediction

Introduction

The potential advantages of synthetic peptide viral vaccines are well known [1,2], including re- duced cost and risk to workers during production, better quality control with no possible contamina- tion with virulent virus, and longer shelf-life. If this technology could be extended to produce an effective rubella vaccine, many of the reported adverse reactions to the current preparation could be excluded. The present vaccine preparations are composed of a live attenuated virus strain,

* Corresponding author. Present address: Department of Pathology, University of Cambridge, Tennis Court Road, Cambridge CB2 lQP, UK.

reported to cause symptoms such as joint pain [3], acute arthritis [4,5] and meningitis [6]. Previous studies have found a whole killed virus vaccine preparation unable to induce high levels of im- munity since agents used in virus inactivation resulted in a loss of antigenicity [7]. A subunit vaccine, therefore, may be a suitable alternative to the present live attenuated form and there is a wealth of evidence that protective immune re- sponses against viral infections can be obtained through vaccination with synthetic peptides [2].

Several studies using monoclonal antibodies have revealed neutralising epitopes in rubella en- velope glycoproteins El and E2 which govern virus infectivity [S-10]. Since El and E2 are ex- pressed in the membranes of infected cells and are the major proteins to which immune re-

SSDZ 0928-8244(94)00083-2

192

sponses are directed during infection [ll] they would be the ideal choice for a synthetic vaccine. Many methods, both practical and theoretical, exist for the elucidation of epitopic sequences within a protein. Since practical methods have been used previously 18-101, a computer algo- rithm for prediction of B cell epitopic sequences was used in this study to provide alternative El and E2 antigenic peptide sequences. Carrier pro- teins, e.g. KLH, tetanus toxoid, PPD, have tradi- tionally been linked to synthetic peptides when used for vaccinations [ll] but several disadvan- tages of this have been reported, including car- rier-induced epitopic suppression [12], cross-reac- tion of antibody with groups in the coupling re- gion [13] and instability and variable immuno- genicity of conjugates depending on the coupling method [14]. Peptide immunogens have been suc- cessfully used without carrier conjugation [15,16], and therefore it was decided to vaccinate using the synthesised peptides in the absence of carrier molecules. Adjuvants have previously been shown to have immunomodulatory activity [17-191. The influence of Freund’s complete adjuvant (FCA) and potassium alum upon antibody isotype re- sponse to peptide vaccination was examined and an indication of the efficacy of peptide vaccina- tion was taken from the induction of neutralising antibody responses.

Materials and Methods

Epitope prediction Sequence analysis of rubella glycoproteins El

and E2 was carried out using the Protean II computer package of Proteus Molecular Design Ltd. This package includes secondary structure prediction methods based on the GOR algorithm [20]. The package uses these predictions together with data on immunogenicity of peptide se- quences and other factors to predict potential B cell epitopic sequences. Four such rubella se- quences were chosen for analysis in this vaccina- tion study. Their sequences are shown in Fig. 1.

Peptide synthesis Peptides were synthesised by solid-phase Fmoc

chemistry [21] using a semi-automated Pepsynthe-

From El:

PXl

309-AHTTSDPWHPPG-320 PX?,

155GDLRAVHHRPVPA-167 PY3

276RGTTPPAYC2az PY4

Fig. 1. Sequences of rubella peptides from envelope proteins El and E2. Amino acids are given in the IUPAC coding system: A = alanine, C = cystine, D = aspartic acid, E = glutamic acid, F = phenylalanine, G = glycine, H = histidine, I = isoleucine, K = lysine, L = leucine, M = methionine, N = asparagine, P = proline, Q = glutamine, R = arginine, S = serine, T = threonine, V = valine, W = tryptophan,

Y = tyrosine.

siser (CRB) with Pepsyn KA resin (Milligen). l-oxo-2-hydroxy-dihydrobenzotriazine esters of Fmoc-protected amino acids were coupled in peptide synthesis grade dimethylformamide (DMF) (Rathburn Chemicals Ltd.) and depro- tected with 20% piperidine (Sigma) in DMF. After synthesis, side-chain protection groups were removed and the peptide cleaved from the resin with scavengers and trifluoroacetic acid (Sigma). Peptides were precipitated in diethylether, redis- solved in distilled water and then freeze-dried. Their purity was assessed by reverse phase HPLC and amino acid analysis (data not shown).

Vaccination Groups of five BALB/c mice were vaccinated

intraperitoneally with 0.1 ml PBS containing 200 pg peptides PXl, PX3, PY3 and PY4, 50 pg commercial rubella antigen (Hoechst), or 0.1 ml culture supernatant containing rubella virus (TCID,, 10-6.75), mixed with the same volume of Freund’s complete adjuvant (FCA) (Sigma). Pep- tide PY4 was adsorbed to aluminium potassium sulphate using a previously optimised method 122,231, immunising doses comprised of 35 mg AlK(SO,), plus 200 pg peptide were injected i.p. A single dose vaccination regime was employed and immune responses were continuously moni-

193

tored in terms of serum antibody by serial tail bleeding.

Serum antibody ELZSA A method adapted from that of Robinson et

al. [24] was used. Briefly, 96-well flat bottomed flexible assay plates (Falcon) were coated with 50 ~1 1 pg ml-’ peptide or a lop4 dilution of rubella antigen (Hoechst) in 0.05 M carbonate-bi- carbonate buffer, pH 9.6 at 4°C overnight. After stringent washing with PBS containing 0.05% Tween,, (PBS-Tween), 50 ~1 of a l/50 dilution of test and naive sera in PBS-Tween was added. The plate was incubated at room temperature for 90 minutes. Plates were washed before addition of 50 ~1 of anti-mouse antibody isotype-alkaline phosphatase conjugalte reagents (Serotec) at the recommended working dilution in PBS-Tween. After a 90-min incubation and further washing, 100 ~1 substrate was added (made up as follows: 1 Sigma 104 Phosphatase substrate tablet per 5 ml diethanolamine buffer, pH 9.8). Plates were incubated at 37°C for 30 min before reading absorbance values at 405 nm on a Dynatech 2000 microplate reader.

virus neutralisation assay Rubella virus strain RA27/3 was cultured in

the adherent cell line RK13 (Flow). Cells were grown in Medium 1’39 (GIBCO) containing 10% foetal calf serum (PCS), 25 mM HEPES, 0.1 mg ml-l r_-glutamine, 100 U ml-’ penicillin, 100 pg

- ’ streptomycin and 1.5% sodium bicarbonate $IBCO) and pasaaged every 3-4 days using trypsin-EDTA (Sigma) treatment. l-day-old cell cultures were used for virus cultivation. Growth medium was replaced with maintenance medium (as above but FCS re:duced to 2%) and inoculated with virus. Flasks were incubated at 37°C with CO, for 5 days before harvesting the virus by a repeated freeze-thaw process. The medium was spun to remove cell debris and the supernatant frozen in liquid nitrogen in 1 ml aliquots [23]. Virus titrations were: determined [25], calculating the dilution giving a 50% tissue culture infectious dose (TCID,,) on RK13 cell cultures in sterile 96-well plates (Nunc). Titres of neutralising anti-

body in serum samples were determined, adapt- ing a previously described method [261 Briefly, 50 pl aliquots of a two-fold dilution series of filter- sterilised serum in PBS were placed into sterile 96-well plates (Nunc). A virus suspension diluted to lOOTCID,, per 50 ~1 aliquot was added to each well. After mixing, the plates were incu- bated at 37°C for 2 h and 100 ~1 of RK13 cell suspension in maintenance medium (1.5 x 10’ cells/ml) was added to each well. Plates were incubated at 37°C with CO, for 5 days before examining the cell monolayers for inhibition of cytopathic effect. The neutralising antibody titre was determined by estimating 50% end-points

]271.

Cytokine ELZSA 200 ~1 aliquots of spleen cell suspensions at

5 X lo6 cells/ml in RPM1 1640 medium (GIBCO) supplemented with 10% FCS, 100 U ml-’ peni- cillin, 100 pg ml-’ streptomycin and 7.5 x 10p5M monothioglycerol (BDH Chemicals Ltd.) were cultured for 48 h in 96-well flat-bottomed tissue culture trays (Nunc) with 5 pg ml-’ Con- canavalin A (ConA) (Sigma) as the peptides were found to be poor T cell stimulators [23]. ConA has been used previously in systems where re- stimulation using antigens was not convenient [28]. IL-5 and IFNy levels in the culture super- natants were measured by sandwich ELISA [28,29]. The monoclonal antibodies TRFK5 and biotinylated TRFK4 (Pharmingen, San Diego, CA) were used for IL-5 assays. R46A2 and bi- otinylated XMG1.2 (Pharmingen) were used to assay for IFNy. Cytokines were quantified by reference to calibrated standards.

Statistical analyses Data for antigen-specific antibody responses of

a given isotype were compared using the Mann Whitney U-test. Data were considered to be sig- nificantly different only where P I 0.05. Assays for nAb were carried out using pooled sera and therefore could not be submitted to statistical analysis. Similarly, pooled spleen cell cultures were used in order to measure T helper cytokine levels.

194

Results

Antibody isotype responses of mice to uaccination with rubella peptides, antigens and virus

Sera from groups of five mice vaccinated previ- ously with peptides PXl, PX3, PY3 or PY4, a commercially available rubella antigen, or culture supernatant containing rubella virus, were tested individually by ELISA for the presence of specific antibody of the isotypes IgGl, IgG2a, IgG2b, IgG3 and IgM (Fig. 2). Synthetic peptide vaccina- tion resulted in antibody responses of mostly IgGl and IgG3. Peptide PY4, however, did not follow this pattern and the predominant isotype pro-

Antibody lsotype

Fig. 2. Antibody isotype responses of mice to vaccination with four rubella peptides. Sera from groups of five mice, vacci- nated 21 days previously with 200 pg peptides PXl, PX3, PY3 or PY4, 50 pg of a commercial rubella antigen, or 0.1 ml culture supernatant containing rubella virus (TCID,, 10-6.75) in Freund’s complete adjuvant, were tested on ELISA plates coated with 1 pg ml-’ of the homologous peptide or antigen for the presence of specific antibody of the isotypes IgGl, IgGZa, IgG2b, IgG3 and IgM. Samples were tested at dilu- tions of l/SO. The differences in mean optical density (OD) (A = 405 nm) between test and naive sera are shown. Error bars represent standard deviations between individual animals

within the groups.

0.5

F a 0.4 z

$ 0.3

E 8 0.2

f 0.1

Y 0

Virus PXl

VE

PX3 PY3

CINATION

PY4

Fig. 3. Reaction of sera raised against four rubella peptides with rubella antigen in ELISA. Sera from groups of five mice, vaccinated 21 days previously with 200 pg peptides PXl, PX3, PY3 or PY4, or 0.1 ml rubella virus supernatant (TCID,, 10-6.75) in Freund’s complete adjuvant, were tested on ELISA plates coated with commercial rubella antigen for the pres- ence of specific antibody. Primary antisera were used at a dilution of l/SO and a polyvalent anti-mouse IgG,A&M- alkaline phosphatase conjugate was the secondary reagent, The differences in mean optical density (OD) (A = 405 nm) between test and naive sera are shown. Error bars represent standard deviations between individual animals within the

groups.

duced was IgG2a. There was a significantly higher level of this isotype in PY4 sera than samples from mice vaccinated with PXl (P = 0.0041, PX3 (P = 0.004) and PY3 (P = 0.004). Vaccination us- ing rubella virus, aIthough the animals could not be infected, also resulted in a predominance of IgG2a. Significantly higher levels of this isotype were produced by virus-vaccination than when rubella antigen was used (P = 0.008). Rubella antigen vaccination resulted in a predominance of IgGl and IgG3.

Sera from each of the peptide-vaccinated groups reacted strongly with a commercial rubella antigenic extract used to coat ELISA plates (Fig. 31, indicating reaction with the native proteins.

Rubella virus neutrahing antibody production as a result of vaccination

Sera from mice vaccinated 21 days previously with peptides PXl, PX3, PY3 or PY4, rubella antigen, or rubella virus, and using FCA as adju- vant, were pooled and tested three times for their capacity to block virus cytopathic effect in vitro, i.e. for nAb (Table 1). Mice vaccinated using virus produced the highest titres whereas those vacci- nated with rubella antigen did not elicit a neutral-

195

ising response. The most effective peptide in in- ducing nAb was PY4 whilst PY3 did not yield any nAb.

T helper responses to vaccination Spleen cells taken from groups of five mice

vaccinated 7 and 14 days previously with either peptide PY4 (the most effective for inducing nAb), rubella antigen, or PBS in FCA as a con- trol, were pooled an.d set in culture for 48 h in the presence of CimA. Supematants were col- lected and tested for the cytokines IL-5 and yIFN, indicative of the T helper subset activated. Vacci- nation with either of these antigens resulted in a higher level of IFNy secretion than IL-5, indicat- ing that the Thl subset was activated in the response (Table 2).

Modulation of the immune response to peptide PY4 by adjuvant

The effect of adj-uvant upon the immune re- sponse to PYCvaccination was investigated. Groups of mice were vaccinated with PY4 in FCA or potassium alum. Sera from these animals

Table 1

Titres of virus-neutralising antibody in murine anti-peptide sera

Vaccination Mean nAb titre

PXl + FCA PX3 + FCA PY3 + FCA PY4 + FCA PBS + FCA PY4 + AlK(SO,), Rubella antigen + FCA Rubella virus Control supernatant

l/26 l/50 Nil l/91 Nil Nil Nil l/114 Nil

Pooled serum samples, from groups of five mice vaccinated 21 days previously with 200 pg peptides PXl, PX3, PY3 or PY4, 50 pg rubella antigen, or 0.1 ml rubella virus (TCID,, 10-6.75) supematant or control supematant in Freund’s complete ad- juvant, or 200 pg PY4 absorbed to AlK(SO,), in a ten-fold dilution series were incub’ated with rubella virus before addi- tion of susceptible cells (RK13). After 5 days incubation at 37”C, 5% CO,, the cell monolayers were scored for evidence of cytopathic effect and the neutralising antibody titre deter- mined. Each sample was assayed on a minimum of two occasions. A positive control human serum contained a mean neutralising antibody titre of l/216.

Table 2

In vitro IL-5 and 1FN-y secretion (U ml-‘) by ConA-stimu- lated spleen cells from vaccinated mice

Priming Day 7-post-vaccination Day 15 post-vaccination antigenic exposure

IL-5 IFNy IL-5 IFNy

in vivo

PY4 0.00 6.44 5.63 27.83 Rubella not done not done 3.91 46.81

antigen PBS 0.00 3.22 not done not done

Spleen cells taken from groups of five mice vaccinated 7 and 14 days previously with either 200 pg peptide PY4 in Freund’s complete adjuvant (FCA), 50 fig commercial rubella antigen in FCA, or with FCA alone, were pooled and set in culture at 5 x lo6 cells per ml for 48 h with 5 Fg ml-’ ConA. Super- natants were collected and tested by ELISA for concentra- tions of IL-S and 1FN-y. Concentrations of IL-5 and IFNy in cell cultures from naive, unvaccinated animals were 0.00 LJ ml-‘.

were analysed for specific IgGl and IgG2a iso- types and also for nAb titres.

Vaccination using FCA appeared to induce a higher and earlier production of IgGl (peak be- tween days 13 and 28) compared with alum (Fig. 4). IgG2a secretion reached similar levels at ap- proximately 28 days post-vaccination. Using alum, however, IgGl and IgG2a reached similar levels around days 28 to 36. No distinct polarity in the production of the two isotypes between the adju- vants was noted, but there was a difference in the timing of responses elicited.

When serum samples taken on day 21 were assayed for nAb, none could be detected in sam- ples from alum-vaccinated animals, whereas FCA-vaccinated mice produced relatively high titres (Table 1).

The algorithm used for this study was very accurate in its prediction of B cell epitopic se- quences since all were immunogenic when used to vaccinate mice, eliciting humoral responses reactive with the native protein. Three of the four successfully induced virus-neutralising responses. The PXl sequence was found within several pub-

196

0.35

0.3

0.25

0.2

0.15

0.1

0.3

0.25

0.2

0.15

0.1

0.05

0 13 20 28 38 42 13 28 36 42

-'13'20'28 36 42

0.2

0.15

0.1

0.05

0 13 20 28 36 42

Days post vaccination

Fig. 4. Comparison of IgGl and IgG2a antibody isotype responses to peptide PY4 injected with Freund’s complete adjuvant and potassium alum. Sera from groups of five mice, vaccinated with 200 pg peptide PY4 in Freund’s complete adjuvant (a,b), or adsorbed to potassium alum (c,d) were tested on ELBA plates coated with 1 pg ml-’ of PY4 for antibody of the isotypes IgGl (a,c) and IgG2a (b,d) on days 13, 20, 28, 36 and 42 post-vaccination. Samples were tested at dilutions of l/50. The differences in mean optical density (OD) (A = 405 nm) between test and naive sera are shown. Error bars represent standard deviations between individual

animals within the groups.

lished rubella haemagglutination and neutralisa- tion epitopes [10,30,31]. These data together indi- cate that algorithms and theoretical methods for epitope prediction may be useful, saving time and expense which are major limitations of most prac- tical methods.

Antibody responses raised against these four small, linear unconjugated peptides appeared to be dominated by the isotype IgG3. This has been shown to be the major IgG surface immunoglobu- lin on murine splenic B cells [34] and the princi- ple subclass directed against carbohydrate anti- gens [35]. The high levels of this isotype may partially be explained by the higher sensitivity of the IgG3 assay compared with other subclasses (data not shown). For example, 1 pg IgG2a in a direct ELISA produced an optical density of

0.581, compared with 0.707 for 1 pg IgG3. Con- cerning other isotypes, the peptides PXl, PX3 and PY3 induced moderate levels of IgGl and low levels of IgG2b and IgM. These data were consistent with IgGl being the major isotype di- rected against soluble protein antigens [321. Pep- tide PY4, however, induced a very high IgG2a response in comparison to the other peptides, with low IgGl levels. It is reported that IgG2a is induced by exposure to viral antigens and infec- tions [32], thus raising the possibility that PY4 may be an appropriate sequence for incorpora- tion into a vaccine formulation.

Antibody isotype responses to vaccination with rubella virus and antigen preparations also re- sulted in a predominance of IgG3, but virus-vac- cination resulted in high IgG2a levels and lower IgGl whilst antigen-vaccination induced opposite profiles. From these data it appeared that the rubella virus preparation induced a similar iso- type profile to peptide PY4, underlining the pos- sible usefulness of this sequence.

Because of the lack of a suitable animal model for rubella, protection studies could not be car- ried out and the best test for potential protective activity of peptide vaccination appeared to be induction of neutralising antibody. Serum sam- ples from virus-vaccinated mice contained the highest titres and PY4 was the most effective peptide. This raised the possibility of a correla- tion between the high IgG2a levels in these sam- ples and higher nAb titres, and IgG2a being the major anti-virus isotype lends support to this hy- pothesis. Further experiments involving fractiona- tion of isotypes in known positive anti-virus or -peptide sera before addition to neutralisation assays would determine whether IgG2a was the isotype, or one of several, responsible for rubella virus-neutralisation.

Immune responses against viral infection are said to include activation of the Thl subset, char- acterised by the secretion of IFNy by T helper cells and the production of IgG2a antibody [33]. The possibility that PY4 induced a Thl-type re- sponse was addressed by assaying culture super- natants from stimulated spleen cells of vaccinated animals for the Thl-associated cytokine IFNy and the Th2 cytokine IL-5 The predominant

197

1FN-y production indlicated a Thl response, and the possibility that PY4 was more efficient in raising nAb due to this appropriate response. The induction of IgGl by the other peptides indicated a possible Th2 type response.

A method to determine whether increased titres of nAb were induced as a result of a Thl- mediated IgG2a response to PYCvaccination was to use adjuvants as immunomodulators of the isotypes elicited. FCA has previously been shown to induce both Thl and Th2 responses 1193, whilst alum induces a Th2 response [17,181 to a given immunogen. When these adjuvants were used with PY4 for vaccination, however, there was no significant polarity in IgGl and IgG2a produc- tion, merely a difference in the timing of these responses. The absence of nAb in sera from alum-vaccinated mice in contrast to the relatively high titres induced by PY4 in FCA was an impor- tant result. This underlines the importance of presentation of immunogens for induction of an appropriate set of responses, particularly if these responses must be protective against a challenge infection.

The nAb responses of animals vaccinated with virus and antigen were puzzling. The antigen preparation induced both IgGl and IgG2a iso- types, stimulated spleen cells secreted IFNy, but no nAb was detected in serum samples. It seems likely that proteins were partially denatured dur- ing the antigen extraction process resulting in destruction of neutralising epitopes, or perhaps the protein fragments forming these epitopes were present in lower pro:portions than in whole virus. Comparisons of antibody specificity could be car- ried out by Western blot analysis.

The success of PY4 in producing relevant im- mune response profiles leading to high nAb titres indicates that synthetic peptide vaccines may be designed with the idea of tailored specific re- sponses. The fact that alum is presently the only licensed adjuvant for human use, when it pro- motes totally inappropriate responses for protec- tion against viral infection, highlights the need for further adjuvant research. Unless synthetic pep- tides can be used wi1.h a good delivery system, the progress of synthetic peptide vaccines may be hindered considerably.

Acknowledgements

This project was funded by the Science and Engineering Research Council and Proteus Molecular Design Ltd.

References

1 Brown, F. (1988) Synthetic peptides as immunogens. Appl. Virol. Res. 1, 93-106.

2 Rowlands, D.J. (1990) Problems and prospects for syn- thetic peptide vaccines. SGM Symposium 45123-142.

3 Benjamin, C.M., Chew, G.C. and Silman, A.J. (1992) Joint and limb symptoms in children after immunisation with measles, mumps, and rubella vaccine. Br. Med. J. 304, 1075-1078.

4 Howson, C.P., Katz, M., Johnston, R.B. and Fineberg, H.V. (1992) Chronic arthritis after rubella vaccination. Clin. Infect. Dis. 15, 307-312.

5 Tingle, A.J., Yang, T., Allen, M., Kettyle, G.D., Larke, R.P.B. and Schulzer, M. (1983) Prospective immunological assessment of arthritis induced by rubella vaccination. Infect. Immun. 40, 22-28.

6 Kihei, M., Nozaki, H., Fukushima, K., Sugishita, K. and Kuriya, N. Comparative study of mumps, mumps meningi- tis and MMR meningitis. Jikeikai Med.J. 38, 361-368.

7 Horstmann, D.M. (1982) Infection of Humans. In: Epi- demiology and Control (Evans, A.S., Ed.), pp. 519-539. Plenum Medical Book Company, New York.

8 Terry, G.M., Ho-Terry, L., Londesborough, P. and Rees, K.R. (1988) Localisation of the rubella El epitopes. Arch. Virol. 98, 189-197.

9 Lozzi, L., Rustici, M., Corti, M., Cusi, M.G., Valensin, P.E., Bracci, L., Santucci, A., Soldani, P., Spreafico, A. and Neri, P. (1990) Structure of rubella El glycoprotein epitopes established by multiple peptide synthesis. Arch. Virol. 110, 271-276.

10 Wolinsky, J.S., Sukholutsky, E., Moore, W.T., Lovett, A., McCarthy, M. and Adame, B. (1993) An antibody- and synthetic peptide-defined rubella virus El glycoprotein neutralization domain. J. Viral. 67, 961-968.

11 Pfaff,E., Mussgay, M., Bohm, H.O., Schulz, G.E. and Schaller, H. (1982) Antibodies against a preselected pep- tide recognise and neutralise foot-and-mouth disease virus. E.M.B.O. J. 1, 869-874.

12 Schulze, M-P., Leclerc, C., Jolivet, M., Audibert, F. and Chedid, L. (1985) Carrier-induced epitopic suppression, a major issue for future synthetic vaccines. J. Immunol. 135, 2319-2322.

13 Edwards, R.J., Singleton, A.M., Boobis, A.R. and Davies, D.S. (1989) Cross-reaction of antibodies to coupling groups used in the production of anti-peptide antibodies. J. Im- munol. Methods 117, 215-220.

14 Briand, J.P., Muller, S. and van Regenmortel, M.H.V. (1985) Synthetic peptides as antigens: pitfalls of conjuga- tion methods. J. Immunol. Methods 78, 59-69.

198

15 DiMarchi, R., Brook, G., Gale, C., Cracknell, V., Doe], T.

and Mowat, N. (1986) Protection of cattle against foot and

mouth disease by a synthetic peptide. Science 232, 639- 641.

16 Francis, M.J., Fry, C.M., Rowlands, D.J., Bittle, J.L.,

Houghten, R.A., Lerner, R.A. and Brown, F. (1987) Im-

mune responses to uncoupled peptides of foot-and-mouth

disease virus. Immunology 61, 1-6.

17 Kenney, J.S., Hughes, B.W., Masada, M.P. and Allison,

A.C. (19891 Influence of adjuvants on the quantity, affin-

ity, isotype and epitope specificity of murine antibodies. J.

Immunol. Methods 121, 157-166.

18 Grun, J.L. and Maurer, P.H. (1989) Different T helper cell

subsets elicited in mice utilizing two different adjuvant

vehicles: the role of endogenous interleukin 1 in prolifera-

tive responses. Cell. Immunol. 121, 134-145.

19 Bomford, R. (1990) Immunomodulation by adjuvants. So-

ciety for General Microbiology Symposium 45, 143-154. 20 Garnier, J., Osguthorpe, D.J. and Robson, B. (1978) Anal-

ysis of the accuracy and implications of simple methods for

predicting the secondary structure of globular proteins. J.

Mol. Biol. 120, 97-120.

21 Atherton, E. and Sheppard, R.C. (1985) Solid phase pep-

tide synthesis using N-fluorenylmethoxycarbonylamino acid

pentafluorophenyl esters. J. Chem. Sot. 165-166.

22 Talmage, D.W. and Dixon, F.J. (19531 The influence of

adjuvants on the elimination of soluble protein antigens

and the associated antibody responses. J. Infect. Dis. 93,

176-180.

23 Robinson, K. (1992) PhD Thesis, University of Manch-

ester.

24 Robinson, K., Bellaby, T. and Wakelin, D. (1994) Vaccina-

tion against the nematode TrichineZlu spiralis in high- and

low-responder mice. Effects of different adjuvants upon

protective immunity and immune responsiveness. Im-

munology 82, 261-267.

25 Grist, N.R., Bell, E.J., Follett. E.A.C. and Urquhart,

G.E.D. (1979) Diagnostic Methods in Clinical Virology,

pp. 81-94. Blackwell Scientific Publications, Oxford.

26 Mougin, B., Bakouche, 0. and Gerlier, D. (1988) Humoral

immune responses elicited in rats by measles viral mem-

brane antigens presented in liposomes and ISCOMs. Vac-

cine 6, 445-495.

27 Reed, L.T. and Muench, H. (1938) A simple method for estimating fifty per cent endpoints. Am. J. Hygiene 27,

493-495.

28 Else, K.J. and Grencis, R.K. (19911 Cellular immune re-

sponses to the murine nematode parasite Trichuris muris.

(I) Differential cytokine production during acute or chronic

infection. Immunology 72, 508-513.

29 Mosmann, T.R. and Fong, T.A.T. (1990) Specific assays

for cytokine production by T cells. J. Immunol. Methods

116, 151-159.

30 Chaye, H., Chong, P., Tripet, B., Brush, B and Gillam, S.

(19921 Localization of the virus neutralizing and hemag-

glutinin epitopes of El glycoprotein of rubella virus. Virol-

ogy 189, 483-492.

31 McCarthy, M., Lovett, A., Kerman, R.H., Overstreet, A.

and Wolinsky, J.S. (1993) Immunodominant T-cell epi-

topes of rubella virus structural proteins defined by syn-

thetic peptides. J. Virol. 67, 673-681.

32 Coutelier, J.P., van der Logt, J.T.M., Heessen, F.W.A.,

Warnier, G. and Van Snick, J. (19871 IgG2a restriction of

murine antibodies elicited by viral infections. J. Exp. Med.

165, 64-69.

33 Mosmann, T.R. and Coffman, R.L. (1989) Heterogeneity

of cytokine secretion patterns and functions of helper T

cells. Adv. Immunol. 46, 111-147.

34 Abney, E.R., Cooper, M.D., Kearney, J.F., Lawton, A.R.

and Parkhouse, R.M.E. (19781 Sequential expression of

immunoglobulin on developing mouse B lymphocytes: a

systemic survey that suggests a model for the generation of

immunoglobulin isotype diversity. J. Immunol. 120, 2041-

2049.

35 Perlmutter, R.M., Hansburg, D., Briles, D.E., Nicolotti,

R.A. and Davie, J.M. (1978) Subclass restriction of murine

anti-carbohydrate antibodies. J. Immunol. 121, 566-572.