NiV is a zoonotic virus that has recently emerged in Malaysia. It has a broad host range and can cause severe respiratory illness and encephalitis with high mortality in humans 
. Despite several previous NiV vaccine studies, there are still no licensed vaccines for human use. Efficacy studies in a non-human primate model are required for the development and approval of a new vaccine or antiviral for use in humans. Recently, African green monkey have been shown to be a highly pathogenic model for NiV infection 
. Pathogenicity of NiV in these monkeys was demonstrated through intratracheal and oral inoculation. When we started to test susceptibility of African green monkeys to NiV, there was no report available. We compared infections resulting from inoculation via
intraperitoneal or intranasal/oral routes. Our findings show that intraperitoneal inoculation induces a more severe manifestation than intranasal and oral inoculation. However, the intranasal/oral route, which mimics the more natural infection route for humans, did cause severe illness in two infected monkeys by 9–14 dpi. Symptoms were consistent with human NiV infection, and the monkeys became moribund, although they eventually recovered. The progression of illness clinically is similar to human cases. Histopathological tests suggested that when administered via
an intraperitoneal route, lymphoid organs including spleen and lung were the main target organs of virus propagation. The cause of death was severe respiratory distress resulting from hemorrhage and edema in the lungs, and monkeys died at 7 dpi, which is well before the viral infection could have advanced to the cerebral region of the brain. Although intranasal/oral inoculation also made monkeys ill, we found no evidence of virus propagation or pathological changes, possibly because the samples were taken at the end of the experiment (24 dpi) after the monkeys had recovered. Intranasal/oral inoculation is more natural route for human infection, and induced symptoms in monkeys similar to those observed in humans. However, we decided to use intraperitoneal route for the NiV challenge after immunization with our recombinant MV vaccine because it caused more severe illness and resulted in an earlier death.
A canarypox virus-based vaccine vector has been shown to be effective as a vaccine against NiV-associated disease in veterinary vaccine 
. The canarypox virus does not replicate in mammalian cells, although it can infect and produce viral proteins. Thus, it is able to eliminate the safety concerns that exist for vaccinia virus vectors. For human use, approaches employing soluble subunit vaccines, virus-like particles, vaccinia virus vectors or complementing defective vesicular stomatitis virus vectors have been explored previously 
. Although they seem promising as vaccines, their efficacy might be problematic because these replication-defective vectors cannot induce long-term immunity.
Live-attenuated measles vaccines have been used since the 1960 s worldwide because they are highly effective and safe. Because MV is an RNA virus, with no DNA intermediates during replication, MV genome does not integrate into host genome. These characteristics make live-attenuated MV vaccines an attractive candidate vector to provide safe and effective immunity against various pathogens. In particular, it induces strong cellular immunity and the effects are long term. For these reasons, many recombinant MVs expressing antigenic proteins of other infectious diseases are under development. It could be argued that the widespread vaccination for measles could result in inactivation of the recombinant MV vector before there is a chance that the NiV G protein can be expressed and induce protective immunity. We examined the antibody responses induced by rMV-Ed-G and rMV-HL-G in MV-seropositive monkeys. Both recombinant MVs could induce specific antibodies; in particular, two inoculation with rMV-Ed-G induced a high titer of a antibodies (1
12800; data not shown). Therefore, it appears that our rMV vaccines are available for people that have been exposed to MV vaccines previously.
In this study, we have demonstrated that recombinant live-attenuated MVs are effective at preventing the onset of symptoms typical of NiV infection. Immunization with recombinant MVs expressing NiV glycoproteins perfectly protected hamsters against a lethal dose upon challenge, although rMV-Ed-G induced NiV-specific IgG antibody level was low in a small number of hamsters. The antibody response is considered to be an essential component of protection against NiV encephalitis 
; however, cellular immunity might play an important role in eradicating NiV infection. We tested two MV vectors, based on our previous experiences where we have observed that rMV-HL-based vaccines sometimes elicit a stronger effect than rMV-Ed-based vaccines. Both rMV-HL-G and rMV-Ed-G induced well protective effect in hamsters against NiV challenge.
The HL strain is isolated from patient and still possesses weak virulence in monkeys. On the other hand, the Edmonston strain was first licensed vaccine in United States in 1963, and further attenuated vaccine derived from the Edmonston strain is widely adopted in the world. Aiming at early practical use, we tested the recombinant Edomonston vaccine in this monkey study. Vaccinated monkeys did not show any symptoms of NiV infection. We used a lower titer (105 TCID50) of NiV for challenge in this experiment, to observe symptoms of infected monkeys for a slightly longer period than those with 106 TCID50. This dose did not induce a lethal pathology, even in unvaccinated individuals. However, histopathological and clinical observations of monkeys indicated that those challenged with NiV did suffer from a severe illness, with unimmunized monkeys found to also have lesions in their brains. The rMV-Ed-G vaccine did completely protect vaccinated monkeys from infection. Further, NiV challenge caused pathological changes in the brain, which has been widely documented in human cases. This observation might be due to slow spreading of the virus in animals challenged with a lower dose.
NiV is highly virulent and has a broad host range, causing respiratory and neurological symptoms that often lead to encephalitis. The rate of mortality in humans range from 40–92% 
. To date, no vaccine for NiV disease has been developed that is both safe and protective in humans. Our recombinant MV-Ed-G vaccine has the potential to elicit long-term immunity against both MV and NiV in children and adults located in endemic areas. Therefore we believe it is an effective vaccine candidate for human use. We were only able to use two monkeys for vaccination in this study, as the costs of non-human primate and spaces for animal experimentation in BSL4 facility were prohibitive. Further studies in greater number of monkeys will be necessary to validate the safety and efficacy of our vaccine candidate.