In light of the recent success of the rinderpest eradication campaign, PPRV would be an excellent candidate for eradication [44, 45] and moves have already been made towards this. The availability of a DIVA vaccine would facilitate PPRV sero-surveillance programmes and speed up the steps leading to disease eradication . In countries newly affected by PPRV, where sporadic outbreaks of disease occur and where the disease is not endemic, a DIVA vaccine would be of value to prevent animal movement restrictions being imposed on countries which cannot prove that animals have been vaccinated and not infected. Since the virus has only 6 genes, all of which are essential for growth, creating a DIVA version of the current live PPRV vaccines would require expressing an extra protein from the viral genome (positively marked vaccine). The alternative is to express one or two viral proteins from an alternative virus vector, thereby eliciting immune protection while not inducing the complete repertoire of antibodies induced following natural infection or vaccination with a live, attenuated PPRV vaccine. This was done successfully for RPV using vaccinia or capripox virus as the vaccine vectors [47–50]. However, these constructs were never used in the field, in part because they did not offer the same duration of protection as the existing vaccine and because the rinderpest eradication campaign was completed without an explicit requirement for a DIVA vaccine. MVA expressing PPRV F and H proteins have been shown to protect goats against subsequent challenge with virulent PPR but two doses of vaccine were given prior to challenge , which would not be practical for a small ruminant vaccination programme. Capripox virus vectors expressing PPRV glycoproteins have been developed [4, 52]. However, in one case, the ability of the vaccine to protect against PPRV was not investigated , while in the other, although protection against PPRV was proven , the vaccine has not been used in the field. The positive aspects of using a recombinant capripox are that it would simultaneously vaccinate against two serious sheep/goat diseases; in addition, capripox-based vaccines would benefit from the intrinsic thermotolerance of poxviruses. However, recombinant capripox vaccines may not be suitable as PPRV DIVA vaccines, as vaccinated animals did not all give good antibody responses, possibly due to pre-existing vector immunity. This is important as it is the comparative antibody response that is likely to be used as the DIVA test, with infected animals having anti-PPRV N and anti-PPRV H antibodies, while the vaccinees will only have anti-PPRV H antibodies. There are existing, well established and validated commercially available cELISAs for both anti-N and anti-H antibodies, making this an attractive DIVA test.
Despite their successful use in several trials as vectors for human vaccines, the FP-based vaccines elicited very poor antibody responses in goats, as well as poor cell-mediated immune responses. The low level of responses in small ruminants may be due to apoptosis of FP virus-infected antigen presenting cells in these animals, as has been shown recently in cattle . Because of the low level of immune response to the PPRV proteins expressed from the FP vectors, we did not pursue these constructs through to challenge, since they would not be useful DIVA vaccines even if effective in protecting the vaccinated animals from PPRV.
Replication-deficient adenovirus-vectored vaccines induce potent CD8+ and CD4+ T-cell responses as well as high antibody responses , and appear to be safe . Furthermore, the Ad vector also acts as an adjuvant . Adjuvant effects in experimental vaccines have been demonstrated by co-expressing cytokines such as IL-2, IL-12 and GMCSF , presenting other options for vaccine formulation. One of the drawbacks of Ad5-vectored vaccines in humans has been that most people have previously been infected with this virus, and the pre-existing antibodies can inhibit the efficacy of the vaccine . However, vaccines based on Ad5 may be suitable for use in livestock since these animals will not have pre-existing immunity to the vector. Large scale production of Ad viruses can be achieved  and, furthermore, Ad viruses can be made more thermostable and efficacious in a range of formulations that further promote stability [59, 60].
Vaccination of goats with Ad-H, alone or with a similar dose of Ad-F, induced levels of H-specific, neutralising antibodies within 2 to 3 weeks that were comparable to those induced by live, attenuated PPRV vaccines [40, 61]. Furthermore, these antibodies were maintained for several months following vaccination. While this work was in progress, similar Ad constructs were reported , which induced a similar level of neutralising antibody as that described in the present study, following a single dose of replication defective Ad-H. In another recent study, replication competent canine adenovirus expressing PPRV H was also found to be effective at eliciting neutralising antibody in goats . Unfortunately, neither of these studies went on to determine the ability of the Ad-vectored vaccines to protect against virulent PPRV. This is important, as the critical elements of the immune response required for protection against PPRV are not yet known. While the current live attenuated PPRV vaccine induces neutralising antibody, and a titre > 1:10 is used as a marker for competency of preparations of such vaccines, attenuated morbillivirus vaccines also induce cell-mediated immunity [64, 65], which may also be important in protection. We have demonstrated that vaccination with Ad-H, or Ad-H and Ad-F, induced a potent effector memory CD8+ response in goats, although the number of H-specific CD8+ IFNγ+ cells had declined to basal levels by 15 weeks post-vaccination. Further studies are therefore needed to determine the effect of Ad-H vaccination on persistence of central memory CD8+ T cells in goats. No detailed studies have been carried out to establish the mechanisms of protection induced by live, attenuated PPRV vaccines. However, studies on rinderpest showed that induction of neutralising antibodies by vaccination with purified viral proteins did not protect against infection , suggesting that it is not possible to deduce protection based on antibody alone. We therefore analysed the ability of Ad-vectored vaccines to protect against infection with virulent PPRV, and demonstrated that a single dose of Ad expressing the PPRV H protein can protect against PPRV challenge up to 4 months after vaccination. Furthermore, vaccinated goats did not appear to excrete infectious virus from the nasopharynx, suggesting that they may not transmit virus to susceptible, unvaccinated animals. This is the first time that an Ad-vectored vaccine has been shown to protect against virulent PPRV challenge, and shows that the immune responses elicited by the replication-defective vaccine are sufficient to protect the vaccinated animal from infection, with a protective response that is sustained for at least 4 months. Longer term studies will be required to determine the duration of PPRV H-specific serum antibodies induced by Ad-H.
In the study by Wang et al. , it was suggested that co-expression of F and H proteins induced higher levels of neutralising antibody than vaccination with Ad expressing either F or H alone. This is a similar finding to that seen in cattle vaccinated with vaccinia virus expressing RPV H, F or both H and F, where it was suggested that the combination of H and F induced stronger protection against RPV . In our studies, expression of H alone (experiment 3) induced neutralising antibody titres at least as high as those seen in animals vaccinated with Ad-F plus Ad-H (experiment 2). This may be because we gave a higher dose of Ad than that used by Wang et al. , and the strength of the response to H alone, coupled with the adjuvant effect of Ad, dominated any co-operative effect of vaccination with Ad H and F together.
This is the first time that the effect of virulent PPRV infection on specific immune cell sub-sets has been analyzed. We have shown that whereas the proportion of circulating WC1+ γ/δ T-cells and CD14+ monocyte/macrophage cells did not change after PPRV infection of control goats, there was a decrease in the proportion of circulating CD4+ cells 4 days after challenge. This decrease may have been due, at least in part, to infection with PPRV, as the proportion of CD4+ cells staining for intracellular H was greater than that of CD8+ cells, 4 days after PPRV challenge. The reduction in CD4+ cells was not observed in any of the Ad-H vaccinated animals. There was a slight increase in the percentage of CD8+ T-cells at 7 days pc in all animals, suggesting induction of CTL responses by PPRV infection. Co-administration of Ad-GMCSF at the time of vaccination appeared to have an effect on infection of CD4+ and CD8+ cells with PPRV. These animals had lower levels of detectable intracellular viral H protein in both the live CD4+ and live CD8+ T cells 4 days after challenge, compared with the other vaccine groups.
The contribution of GMCSF and IL-2 in boosting immune responses to Ad vaccination is not clear from this study. However, the results from the 2nd experiment suggested that the combination of Ad-IL-2 and Ad-GMCSF induced higher H-specific serum antibodies and a greater H-specific CD8+ IFNγ+ response compared with Ad-F and Ad-H alone. The level of N-specific antibodies after challenge with PPRV was significantly lower in the two vaccine groups that received Ad-IL-2 compared with the other groups of goats, suggesting that there was less replication of the challenge virus in these animals and, therefore, that co-administration of Ad-IL-2 induced a more effective protective immune response, even though a significant effect on T cell responses, H-specific or neutralising antibody levels was not seen. It will be interesting to determine if an adjuvant effect of the co-expressed cytokines is more obvious at lower doses of Ad-H/Ad-F. If such studies demonstrate that a co-administered Ad-vectored cytokine has a dose sparing effect on an Ad-vectored PPRV vaccine, then it may be possible to construct a recombinant Ad, which can have an insert of ~7.5 kb, expressing both PPRV H and cytokine.
In conclusion, we have demonstrated that a single vaccination with a recombinant Ad expressing the PPRV H protein induced PPRV-specific neutralising antibodies, primed CD8+ T cells, was safe, and completely protected goats against PPR for at least 4 months.