We are developing human parainfluenza viruses (HPIVs) and avian paramyxoviruses (APMVs) as vaccine vectors for human use against highly pathogenic emerging viruses, using HPIV3, B/HPIV3, and the APMV Newcastle disease virus (NDV) as proof of principle. This strategy takes advantage of the natural respiratory tract tropism of HPIVs and NDV to provide respiratory administration and stimulate strong systemic immune responses as well as local mucosal immunity that is important for restricting pathogens that infect and are spread via the respiratory tract and conjunctiva. With PIV3 vectors, we previously performed extensive pre-clinical studies with constructs expressing the SARS S and Ebola GP glycoproteins. In these first generation (1st gen) constructs, the PIV vector contained the complete viral genome with the foreign antigen expressed from an added gene. When given by the combined intranasal (IN) and intratracheal (IT) routes, these constructs were substantially immunogenic and protective in rodents and non-human primates, even in animals previously infected with the empty PIV3 vector. However, immunogenicity depended on IT delivery of vaccine: IN delivery alone was insufficient. This suggested that vector expression beyond the upper respiratory tract was necessary for immunogenicity. However, IT delivery in humans would not be feasible. Therefore, we used a 1st gen construct consisting of HPIV3 expressing Ebola GP, called HPIV3/EboGP, to explore aerosolized delivery in rhesus macaques. The aerosol route is one that is feasible for use in humans, and indeed delivery devices exist and have been in human use. The aerosol route was generally more immunogenic and protective than the combined IN/IT route. This induced generally higher serum and mucosal EBOV-specific IgG, IgA, and neutralizing antibody titers, as well as EBOV-specific cellular responses in the lungs, including polyfunctional CD8+ T cells and a subset that expressed CD103 (resident T cell marker), and CD4+ T helper cells that were predominately type 1. In addition, the HPIV3/EboGP vaccine produced more robust cell-mediated and humoral immune responses than a systemic alphavirus vaccine delivered in parallel. One aerosol dose of HPIV3/EboGP conferred 100% protection to macaques against Ebola challenge. We presently are performing (with JHU clinical collaborators) an open label vaccine trial to determine the safety, tolerability and immunogenicity of HPIV3/EboGP (the 1st gen construct) delivered IN in healthy adults in an in-patient setting (NCT02564575). This was designed as a safety study prior to a planned second study to evaluate aerosol delivery. In the first study, an initial cohort of subjects will receive 2 doses, at a 4- to 8-week interval, of 106 PFU of vaccine. Subsequently, a second cohort will receive 2 doses, at a 4- to 8-week interval, of 107 PFU of vaccine. We also previously developed a second generation (2nd gen) version of HPIV3/EboGP in which the HPIV3 HN and F genes were deleted. The resulting construct, called HPIV3/delHNF-EboGP, has Ebola GP as the sole viral surface glycoprotein. This construct is replication-competent, and was highly attenuated and yet immunogenic in rodents. This 2nd gen construct is presently being evaluated in non-human primates by the IN/IT and aerosol routes. We have prepared clinical trial material in anticipation of clinical trials. APMVs such as NDV are antigenically distinct from common human pathogens and thus should be unaffected by the infection history of the human recipient. They also are potent inducers of interferons, which are thought to have an adjuvant effect. We previously showed that NDV is very highly attenuated in non-human primates, and was immunogenic and protective as a vaccine vector for SARS and HPAIV glycoproteins. NDV strains naturally exhibit a spectrum of virulence, from low-virulence strains that can be used as live poultry vaccines to high-virulence strains that can cause epidemics with high mortality in poultry. Note that NDV virulence is defined by classic pathogenicity tests of mean embryo death time in embryonated chicken eggs and intracerebral pathogenicity index in 1-day-old chicks. In mammalian hosts, NDV strains that have intermediate-virulence (in chickens) tend to replicate better and be more immunogenic than strains with low-virulence (in chickens). However, intermediate-virulence strains typically have (i) a polybasic cleavage site in the F protein and (ii) relatively high scores on standard chicken pathogenicity tests, both of which define these strains (along with high-virulence strains) as Select Agents, which greatly complicates their use and raises agricultural concerns. The elimination of the polybasic cleavage site and a reduction in virulence scores would remove their Select Agent status and eliminate agricultural concerns. Thus, for example, our collaborator Dr. Samal (University of MD) modified the intermediate-virulence strain NDV-BC by swapping its polybasic cleavage site in the F protein with a non-polybasic cleavage site from the low-virulence LaSota (Las) strain (note that all work with Select Agents for this project was done under appropriate containment at UMD, and no work with Select Agents was performed at NIH). This also resulted in a dramatic reduction in virulence score, and thus this strain (NDV-BC-Las-Fc) is no longer a Select Agent. However, this also resulted in a reduction in its replication and immunogenicity. Dr. Samal, in a collaboration, has been working to increase replication and immunogenicity while retaining low virulence, such as by swapping other domains in the F and HN glycoproteins between NDV-BC-Las-Fc and high-virulence strains. In some cases, this has increased replication and immunogenicity while retaining low virulence. This initially involved swapping large domains of the F and HN proteins, followed by down-selection to identify relevant smaller domains and individual amino acid residues. For example, four amino acid assignments in the F1 subunit (341S, 384I, 385A, and 386L) were found to confer increased replication and fusogenicity without a substantial increase in virulence, while another assignment (403D) had the opposite effect. Thus, we have identified several derivatives of interest that have given preliminary evidence of improved performance and will be evaluated further as vaccine vectors, including in non-human primates.
Showing the most recent 10 out of 52 publications
