The effect of the TLR9 ligand CpG-oligodeoxynucleotide on the protective immune response to alcelaphine herpesvirus-1-mediated malignant catarrhal fever in cattle

We wished to determine the effect of of CpG ODN adjuvant on the magnitude and duration of protective immunity against alcelaphine herpesvirus-1 (AlHV-1) malignant catarrhal fever (MCF), a fatal lymphoproliferative disease of cattle. Immunity was associated with a mucosal barrier of virus-neutralising antibody. The results showed that CpG ODN included either with emulsigen adjuvant and attenuated AlHV-1 (atAlHV-1) or alone with atAlHV-1 did not affect the overall protection from clinical disease or duration of immunity achieved using emulsigen and atAlHV-1. This is in contrast to other similar studies in cattle with BoHV-1 or cattle and pigs with various other immunogens. In addition to this, several other novel observations were made, not reported previously. Firstly, we were able to statistically verify that vaccine protection against MCF was associated with virus-neutralising antibodies (nAbs) in nasal secretions but was not associated with antibodies in blood plasma, nor with total virus-specific antibody (tAb) titres in either nasal secretions or blood plasma. Furthermore, CpG ODN alone as adjuvant did not support the generation of virus-neutralising antibodies. Secondly, there was a significant boost in tAb in animals with MCF comparing titres before and after challenge. This was not seen with protected animals. Finally, there was a strong IFN-γ response in animals with emulsigen and atAlHV-1 immunisation, as measured by IFN-γ secreting PBMC in culture (and a lack of IL-4) that was not affected by the inclusion of CpG ODN. This suggests that nAbs at the oro-nasal-pharyngeal region are important in protection against AlHV-1 MCF.     

A new inactivated BVDV genotype I and II vaccine. An immunisation and challenge study with BVDV genotype I

An inactivated vaccine containing BVDV I and II strains (PT810; BVDV I, and 890; BVDV II) and using different adjuvants and antigen dosages was tested in a cattle challenge model. Groups of six healthy, seronegative cattle were vaccinated twice with a low dose (10(6.6) TCID(50) PT810 and 10(7.2) TCID(50) 890) vaccine with the adjuvant Bay R1005 or a high dose (10(7.8) TCID(50) PT810 and 10(8. 2) TCID(50) 890) vaccine with two different adjuvants (Bay R1005 or Polygen). Thirty-eight days after the second vaccination, immunised animals (n=18) and non-vaccinated control animals (n=3) were challenged intranasally with 10(6) TCID(50) BVDV strain PT810. For a period of 16 days, virus was isolated from blood leukocytes and nasal swabs, and neutralising antibody titres were determined.The induction of antibodies following immunisation was strongly dependent on the antigen dosage in the vaccine. The high dose formulation induced high serum neutralising antibody titres against both genotypes of up to 32000 after the second immunisation. Animals with neutralising antibody titres >512 (n=14) did not show any marked leukopenia after challenge and only very little or no virus could be isolated from blood leukocytes and/or nasal swabs when compared to control cattle. Furthermore, some of these animals did not show any boost of neutralising or even NS3-specific antibodies, which renders viral replication unlikely and thus would prevent infection of the fetus. Both adjuvants (Bay R1005 or Polygen) were similarly efficient and induced nearly identical antibody responses. In contrast, four of the six low dosage vaccinates had a marked leukopenia and viraemia as well as detectable nasal virus shedding for several days.We conclude that the selected strains and the system of vaccine preparation with high BVDV antigen dosages and highly efficient new adjuvants provide an effective means of protection against BVDV I infections. Investigations to demonstrate the protection against BVDV II infections, the duration of immunity and the ability of fetal protection by using the high dose vaccine in a fetal challenge model will follow.     

Comparative immunogenicity of two bivalent botulinum vaccines

OBJECTIVE: To compare the ability of a new single-dose botulinum vaccine containing a non-mineral oil adjuvant with a single dose of a conventional botulinum vaccine product to produce antibody to Clostridium botulinum types C and D in cattle in Northern Australia. DESIGN AND PROCEDURE: One hundred and fifty Brahman steer weaners were randomly divided into two groups receiving either a single dose of CSL Bivalent Botulinum vaccine or Websters Singvac. Blood samples were collected at 0, 8 and 24 weeks and tested by antibody ELISA. The final samples were also tested by the toxin neutralisation test, to test titres of neutralising antibody. RESULTS: Six months after inoculation, cattle vaccinated with Websters Singvac had ELISA antibody response twice that of CSL conventional product. However, this difference was only evident for neutralising antibody to type C botulinum toxin. Both products produced similar titres of type D neutralising antibody after a single dose. CONCLUSION: Websters' Singvac produces a greater neutralising antibody response to type C botulism upon single inoculation than a conventional vaccine. The product produces an equivalent neutralising antibody response to type D.     

A live bovine herpesvirus-1 marker vaccine is not shed after intramuscular vaccination

It should be established, whether animals vaccinated intramuscularly (IM) with a live Bovine herpesvirus type 1 (BHV-1) marker vaccine become viremic and/or excrete vaccine virus with nasal discharge. Five cattle, seronegative for BHV-1, were vaccinated with an overdose of the vaccine (Bovilis IBR marker live) via the IM route. Nasal swabs and blood samples were taken at regular intervals and tested for BHV-1 in a virus infectivity assay. In addition, a polymerase chain reaction (PCR) specific for BHV-1 DNA was performed on the blood samples. BHV-1 neutralizing antibody titres were determined in the sera taken prior to the vaccination and four weeks after immunisation. AIl animals were successfully vaccinated as judged by the development of BHV-1 neutralising antibodies. However, all nasal swab samples were tested negative for vaccine virus, and all blood samples were found negative for BHV-1 vaccine virus and BHV-1 specific DNA. From these data it can be concluded that the vaccine virus was not excreted with nasal discharge after IM vaccination and that the vaccinated animals did not have a detectable viremia. Therefore, it is recommended to apply the tested BHV-1 marker live vaccine by the IM route in situations where it is undesirable that the vaccine virus is excreted.     

Vaccination of calves against bovine herpesvirus-1: assessment of the protective value of eight vaccines

Eight separate, but related experiments, were carried out in which groups of six calves were vaccinated with one of eight commercial vaccines. In each experiment the vaccinated calves were subsequently exposed to three calves infected with virulent bovine herpesvirus-1 (BHV-1). In each experiment, all infected donor calves developed a typical severe infectious bovine rhinotracheitis (IBR) infection and excreted virus in their nasal secretions of up to 10(8.00) TCID50/0.1 ml. One live BHV-1 gE-negative vaccine (A) and three modified live vaccines (B, C, D), administered intranasally, all protected against clinical disease. The calves vaccinated with one vaccine (C) also did not excrete virus in the nasal secretions, whereas the calves protected by vaccines A, B and D excreted virus in their nasal secretions but at low titres (10(0.66)-10(1.24) TCID50/0.1 ml). A fourth modified live vaccine (E), given intramuscularly, failed to prevent mild clinical disease in the calves which also excreted virus in the nasal secretions at titre of 10(1.00) TCID50/0.1 ml. An analogous result was given by the calves vaccinated with either of the two inactivated vaccines (F and G) or with a BHV-1 subunit vaccine (H). All calves developed mild clinical signs and excreted virus at titres of 10(2.20)-10(3.12) TCID50/0.1 ml. Calves vaccinated with C vaccine were subsequently given dexamethasone, following which virus was recovered from their nasal secretions. The virus isolates did not cause disease when calves were infected and appeared to be closely related to the vaccine strain.     

Impact of field vaccination with a Theileria annulata schizont cell culture vaccine on the epidemiology of tropical theileriosis

Tropical theileriosis, caused by Theileria annulata, is an important tick-borne disease of cattle. A cell culture attenuated vaccine has been developed in our laboratory by long-term in vitro propagation of the schizont stage of the parasite. A longitudinal study was conducted at selected farms housing indigenous, cross-bred and exotic animals to investigate the effect of vaccination on the epidemiology of the disease. A total of 120 animals in 4 age groups were vaccinated with the vaccine before the onset of disease season. An equal number of age-matched animals were kept as controls at the same sites. Animals were monitored for 14 months at monthly intervals. The 97.5% vaccinated animals showed a rise in antibody titres 1 month post-vaccination, as determined by single dilution ELISA. The 78.3% of non-vaccinated animals became sero-positive over the period of observation. Mean antibody titres were significantly higher in vaccinated than non-vaccinated animals. Cross-bred animals showed higher antibody titres followed by exotic and indigenous animals in both the vaccinated and non-vaccinated groups. However, the antibody titres in animals of different ages were similar. The 36.7% vaccinated and 64.2% non-vaccinated animals became carriers (<0.5% piroplasms in erythrocytes) during the observation period. Clinical cases of theileriosis were recorded only in the non-vaccinated group suggesting that vaccinated animals were sufficiently immune to withstand field tick challenge for at least 14 months.     

Testing the effectiveness of foot-and-mouth disease vaccines: the relationship between test infection results and corresponding neutralization titers of vaccinated cattle

In the course of vaccine controls, the potency of 25 foot-and-mouth disease (FMD) vaccines was tested quantitatively in parallel in cattle using the intradermal infection and the determination of the SN titres. More than 95% of the vaccinated cattle with SN titres of greater than 1:20 were protected from generalized FMD, regardless of the virus type tested. 61.5% of the vaccinated cattle with SN titres less than or equal to 1:20 were not protected and developed generalized FMD. Comparison of the PD50 values calculated from the results of the intradermal infection and the corresponding SN titres (minimum protection titre greater than 1:20) showed that the results were in complete agreement in 56% of the tested vaccines. In a further 32% of vaccines, the PD50 calculated from the SN titre was slightly below that for the intradermal infection, in the remaining 12% it was somewhat above. The possibility of using the minimum titre determination for testing a vaccine and the significance of this titre as an expression of protection by vaccination are discussed.     

Vaccination of Tunisian dogs with the lyophilised SAG2 oral rabies vaccine incorporated into the DBL2 dog bait

The protective effect of the lyophilised SAG2 oral vaccine bait DBL2, already demonstrated on laboratory dogs, needed to be verified on common Tunisian dogs. Seven Tunisian dogs consumed totally or partially one DBL2 bait containing 10(8.3) TCID50 of the highly attenuated rabies vaccine strain, SAG2. Five of the seven vaccinated animals survived a challenge administered 33 days later with a Tunisian canine street rabies virus to which five of the six controls that were not vaccinated and had no specific antibodies succumbed. The partial or total consumption of a single DBL2 bait thus conferred a protective immune response similar to that observed in laboratory dogs to dogs of poor health status. The sero-antibody response was, however, weak: only two vaccinated dogs exhibited a significant neutralising antibody response after vaccination and before the challenge, and four after the challenge.     

Analysis of the quality of protection induced by a porcine influenza A vaccine to challenge with an H3N2 virus

Antigenic drift of swine influenza A (H3N2) viruses away from the human A/Port Chalmers/1/73 (H3N2) strain, used in current commercial swine influenza vaccines, has been demonstrated in The Netherlands and Belgium. Therefore, replacement of this human strain by a more recent swine H3N2 isolate has to be considered. In this study, the efficacy of a current commercial swine influenza vaccine to protect pigs against a recent Dutch field strain (A/Sw/Oedenrode/96) was assessed. To evaluate the level of protection induced by the vaccine it was compared with the optimal protection induced by a previous homologous infection. Development of fever, virus excretion, and viral transmission to unchallenged group mates were determined to evaluate protection. The vaccine appeared efficacious in the experiment because it was able to prevent fever and virus transmission to the unchallenged group mates. Nevertheless, the protection conferred by the vaccine was sub-optimal because vaccinated pigs excreted influenza virus for a short period of time after challenge, whereas naturally immune pigs appeared completely protected. The immune response was monitored, to investigate why the vaccine conferred a sub-optimal protection. The haemagglutination inhibiting and virus neutralising antibody responses in sera, the nucleoprotein-specific IgM, IgG, and IgA antibody responses in sera and nasal secretions and the influenza-specific lymphoproliferation responses in the blood were studied. Vaccinated pigs developed the same or higher serum haemagglutination inhibiting, virus neutralising, and nucleoprotein-specific IgG antibody titres as infected pigs but lower nasal IgA titres and lymphoproliferation responses. The lower mucosal and cell-mediated immune responses may explain why protection after vaccination was sub-optimal.     

Immunization of cattle against Aujeszky's disease with inactivated adjuvant vaccine

The experiments with sheep and young cattle were carried out to test the immunizing efficacy of inactivated adjuvant vaccine against Aujeszky's disease. The vaccine application at doses of 1 ml and 2 ml to lambs at the age of eight to ten months caused the neutralizing antibody production with a significant rise of titres after revaccination. A survival of infection induced with a dose of 10(5.5) TKID50 of virulent virus was recorded in 62.5% of once vaccinated animals and in 87.5% of twice vaccinated animals. When applying different doses of vaccines (from 1 to 10 ml) to young cattle, the antibody reaction level was directly dependent on the inoculum quantity. The double inoculation of animals with vaccines of 2 ml and 5 ml caused the neutralizing antibody production at titres of 1:35, or 1:46. The animals, immunized with the live or inactivated IBR-vaccine possessing high antibody titres against IBR-virus, reacted upon the vaccination with inactivated Aujeszky's vaccine anamnestically, by early production of antibodies in high titres. Metaphylactic vaccination (2 ml of vaccine) of cattle in herds with an acute course days, however earlier during five days from the revaccination when it was carried out in seven days following the first vaccination.     

Immune responses of sheep to louping-ill virus vaccine

The immune responses of sheep to single and double doses of commercially available louping-ill virus vaccine were examined. The susceptibility to challenge of sheep which had been vaccinated but showed a poor response was also investigated. Two injections of vaccine were required to provoke an adequate antibody response and maximum titres were obtained when there was an interval of two to eight weeks between injections. After challenge, viraemia could not be detected in animals with an antibody titre of 20 although increase in the concentration of humoral antibodies indicated that infection had occurred. Vaccinated but seronegative sheep and vaccinated animals with an antibody titre of 10 were also clinically resistant to the challenge, although circulation of virus was demonstrated. That vaccination had sensitised those animals to viral antigen was evident from the reduced viraemias, the early rise in humoral antibody titres and subsequent protection afforded compared to unvaccinated control animals. Thus, animals with minimal antibody titres after vaccination are protected, but it is recommended that vaccines eliciting the highest possible antibody responses will be the most useful under field conditions.     

Does multiple oral vaccination of wild boar against classical swine fever (CSF) have a positive influence on the immunity?

We studied the efficacy of multiple vaccinations of wild boar against classical swine fever (CSF) using a C-strain vaccine. The study consisted of two experiments. In the first experiment, 7 to 8 months old animals were vaccinated either three or four times at an interval of 7 days or twice at an interval of 14 or 28 days. In the second experiment, the efficacy of oral immunisation in young boars (3 months old) was examined after fivefold vaccination at intervals of 14 or 28 days. Independently of the immunisation scheme all wild boar developed neutralising antibodies. An evaluation of the antibody titres 28 days after the initial vaccine application showed that single vaccination and triple immunisation at an interval of 7 days induced the highest antibody titres (X > or = 1/80). In multiple vaccinated young boars (vaccinated at intervals of 14 or 28 days) the third vaccination led to a slight reduction or to an only moderate increase of the antibody titre. In a challenge study after the fifth vaccination all wild boar were protected (no viraemia, no virus excretion, no post-mortem virus detection in organs). This was confirmed by the fact that sentinel animals were not affected. Although other immunisation schemes also were effective, booster vaccination at an interval of 28 days is recommended as basic procedure for eradication of CSF in wild boar. Triple vaccination might also be used at the beginning of the control measures.     

A bovine ephemeral fever vaccine incorporating adjuvant Quil A: a comparative study using adjuvants Quil A, aluminium hydroxide gel and dextran sulphate

Various vaccines containing the 919 strain of ephemeral fever virus were evaluated in experimental calves and in commercial cattle. The vaccine virus was mixed with one of the adjuvants, Quil A (a saponin derivative), aluminium hydroxide gel, dextran sulphate or combinations of these. The response of experimental calves was evaluated by measuring the production of neutralising antibodies and by resistance to challenge with virulent virus; the response of commercial cattle was judged only by the production of neutralising antibody. Twelve calves given two doses of vaccine containing Quil A produced neutralising antibodies to bovine ephemeral fever virus and all were resistant to challenge with virulent virus given 28 to 76 days after the second vaccination. The vaccine given in three of these calves also contained aluminium hydroxide gel. Six of eight unvaccinated control calves succumbed to experimental challenge. In commercial cattle (17 to 26 animals per group) the serological response after two doses of vaccine containing Quil A or Quil A and dextran sulphate was significantly better than that after vaccines containing only dextran sulphate or after vaccines containing combinations of aluminium hydroxide gel and Quil A. The adjuvant Quil A alone was tested in cattle and shown to produce a transient soft swelling at the injection site as well as a rise in rectal temperature of greater than 1 degree C one day after inoculation. At least 99.99 per cent of viral infectivity was destroyed when the vaccine was mixed with Quil A, suggesting that live virus may not be essential in the immunogenicity of the vaccine. This vaccine overcame two of the problems associated with previous attenuated vaccines tested in Australia; the necessity for adjuvant and virus to be mixed immediately before use and the large volume of the vaccine.     

Measurement of antibody in serum and genital fluids of bulls by ELISA after vaccination and challenge with Tritrichomonas foetus

An enzyme-linked immunosorbent assay (ELISA) was developed for detecting antibody to Tritrichomonas foetus using both whole cell antigen (WCA) and membrane protein antigen (MPA). The test was used to detect specific antibody in serum, preputial washings and seminal plasma samples from 7 adult bulls which were vaccinated subcutaneously on 3 occasions with a membrane protein vaccine against T. foetus var brisbane in an oil adjuvant, and from 4 unvaccinated control animals. One month after administration of the third dose of vaccine, vaccinated and control bulls were repeatedly challenged with the live vaccine strain of the T. foetus. A steady increase in serum antibody titre was detected after each inoculation of vaccine when both antigens were used in the ELISA. However, MPA was more sensitive. After challenge, vaccinated bulls developed an increased titre. No specific antibody was detected in control bulls, except in one bull after challenge in which seroconversion was detected. The serum antibody titres of both groups of animals were also measured with the microagglutination test which proved less sensitive than the ELISA. Antibody titres to both antigens, although lower than in serum, were detected in the seminal plasma of vaccinated animals. The control bulls remained non-responsive. No antibody was detected by ELISA in preputial washings from either control or vaccinated bulls prior to challenge. Post-challenge, some of the vaccinated bulls were responsive with both antigens whereas the control bulls remained negative.     

A CELL CULTURE VACCINE AGAINST BOVINE EPHEMERAL FEVER

SUMMARY A vaccine was prepared from cell culture fluids harvested from the twelfth passage of the 919 strain of bovine ephemeral fever (BEF) virus in Vero cell cultures. Cattle were vaccinated subcutaneously with various combinations of strain 919 virus and adjuvants. Neutralising antibodies were assayed at various times after vaccination and some cattle were challenged by intravenous inoculation with the virulent 417WBC strain of BEF virus. Strain 919 virus of the third and twelfth passage levels in Vero cells produced neither fever, clinical illness nor detectable viraemia in 5 calves inoculated intravenously. Nor could viraemia be detected in 5 heifers receiving vaccine subcutaneously. When the vaccine was administered mixed with aluminium hydroxide adjuvant, the production of neutralising antibodies increased with an increase in the volume of vaccine from 2.5 ml to 10 ml and the response to 2 injections was significantly better than the response to a single injection. The neutralising antibody response was decreased when vaccine was diluted in phosphate buffered saline. The neutralising antibody response following 2 subcutaneous vaccinations with strain 919 virus mixed with aluminium hydroxide adjuvant was higher than that following intravenous inoculation with virulent virus. The vaccine-induced antibodies persisted for at least 12 months, and revaccination at this time led to an increase in the titre of neutralising antibody. Antibodies induced by a single subcutaneous administration of strain 919 virus mixed with Freund's complete adjuvant persisted for at least 40 weeks; those induced by vaccine containing Freund's incomplete adjuvant had virtually disappeared within 16 weeks. All these calves responded to vaccination with aluminium hydroxide-containing vaccine with increases in levels of neutralising antibodies. Of 26 vaccinated calves challenged with virulent BEF virus, 24 remained clinically normal. Two developed brief periods of pyrexia on the seventh day after challenge, but no other clinical signs. One of these calves had a viraemia that was demonstrated only by intravenous inoculation of a susceptible calf. The remaining calf had no detectable viraemia. All of 7 unvaccinated calves developed severe clinical BEF within 5 days of challenge. No disease attributable to the 919 virus occurred in 24 vaccinated pregnant heifers or their newborn calves.     

Protective efficacy of a live attenuated Mycoplasma hyopneumoniae vaccine with an ISCOM-matrix adjuvant in pigs

An attenuated Mycoplasma hyopneumoniae vaccine that requires intrathoracic administration is commercially available for use against mycoplasmal pneumonia in China. Given the limitations of such a route of administration, this study was undertaken to assess the capacity of an ISCOM-matrix adjuvant to enhance immunogenicity following intramuscular use. Immune responses in pigs following vaccination and subsequent intra-tracheal bacterial inoculation were examined using lymphocyte proliferation, serology and mucosal IgA in both nasal and saliva swabs.Vaccination induced clear lymphocyte proliferation, but only slight serum antibody responses although these were significantly increased following experimental infection. Mucosal IgA was not detected in either nasal or salivary secretions. Following bacterial challenge, animals vaccinated with the adjuvant-containing live vaccine exhibited less severe pulmonary lesions (median score 3.67) than unvaccinated pigs (median score 13.58). The degree of ciliary loss on the respiratory tract surface was reduced in vaccinated pigs compared with experimentally infected controls. The findings indicated that the adjuvant vaccine administered IM provided protection against experimentally induced mycoplasmal pneumonia and could have commercial potential.     

Seroconversion in captive African wild dogs (Lycaon pictus) following administration of a chicken head bait/SAG-2 oral rabies vaccine combination

This study determined the proportion of captive juvenile and adult African wild dogs (Lycaon pictus) that developed protective titres of rabies neutralising antibodies following ingestion of a chicken head bait/SAG-2 oral rabies vaccine combination. A single chicken head containing 1.8 ml of SAG-2 vaccine (10(8.0) TCID50/ml) in a plastic blister was fed to each of eight adult and three juvenile wild dogs. Bait ingestion resulted in a significant rise in serum neutralising antibody titres. Overall seroconversion rate was eight out of 11 (72.7%), and all the puppies and five out of eight (62.5%) adults showed potentially protective levels of antibodies on day 31. The mean post-vaccination neutralising antibody titre was within the range reported to be protective against challenge with virulent rabies virus in other species.     

The effect of dose, route and virulence of bovine herpesvirus 1 vaccine on experimental respiratory disease in cattle.

Three experiments were conducted with calves in which, following intramuscular or intranasal vaccination with virulent or attenuated bovine herpesvirus 1, calves were protected against bovine herpesvirus 1 -- Pasteurella haemolytica challenge. Calves receiving low doses of vaccine had lower levels of antibody and greater evidence of virus replication upon challenge than those receiving higher doses. In contrast 11/13 unvaccinated controls had fibrino-purulent pneumonia following challenge. The immune response developed later in younger calves and those given low doses of vaccine. Neutralizing antibodies to bovine herpes-virus 1 were not found in nasal secretions, but were present in serum seven days after vaccination. Bovine herpesvirus 1 was isolated before challenge from nasal secretions of calves vaccinated intranasally or intramuscularly with virulent virus but not those vaccinated intramuscularly with vaccine virus. It was concluded that both routes of vaccination with either virulent or attenuated bovine herpesvirus 1 provided protection from challenge with homologous or heterologous bovine herpesvirus 1 and that live vaccines should contain at least 10(3) plaque forming units/dose for effective immunization.     

Maturation of functional antibody affinity in animals immunised with synthetic foot-and-mouth disease virus.

A good correlation exists between specific neutralising antibody titre and protection against challenge with foot-and-mouth disease virus (FMDV) in infected or virus-vaccinated cattle, but not in the case of animals immunised with synthetic FMDV peptides. Therefore, mechanisms other than simple neutralisation are likely to be important in vivo. Antibody affinity may influence the protective capacity of sera from immunised animals and experiments were carried out to measure the functional affinity for synthetic FMDV peptide of sera from guinea pigs and cattle given various synthetic vaccines. In guinea pigs given a single dose of synthetic vaccine, antibody affinity increased with time after immunisation. In cattle, however, administration of a second dose of peptide 21 days after the first markedly retarded the process of affinity maturation. For guinea pig sera of equivalent neutralising activity, those of higher functional affinity had higher protective indices than those of lower functional affinity. Knowledge of the importance of antibody affinity in protection against FMD is important for an improved understanding of the mechanisms of protection and for the design of novel vaccines.