During a global influenza pandemic, the vaccine requirements of developing countries can surpass their supply capabilities, if these exist at all, compelling them to rely on developed countries for stocks that may not be available in time. There is thus a need for developing countries in general to produce their own pandemic and possibly seasonal influenza vaccines. Here we describe the development of a plant-based platform for producing influenza vaccines locally, in South Africa. Plant-produced influenza vaccine candidates are quicker to develop and potentially cheaper than egg-produced influenza vaccines, and their production can be rapidly upscaled. In this study, we investigated the feasibility of producing a vaccine to the highly pathogenic avian influenza A subtype H5N1 virus, the most generally virulent influenza virus identified to date. Two variants of the haemagglutinin (HA) surface glycoprotein gene were synthesised for optimum expression in plants: these were the full-length HA gene (H5) and a truncated form lacking the transmembrane domain (H5tr). The genes were cloned into a panel of Agrobacterium tumefaciens binary plant expression vectors in order to test HA accumulation in different cell compartments. The constructs were transiently expressed in tobacco by means of agroinfiltration. Stable transgenic tobacco plants were also generated to provide seed for stable storage of the material as a pre-pandemic strategy. For both transient and transgenic expression systems the highest accumulation of full-length H5 protein occurred in the apoplastic spaces, while the highest accumulation of H5tr was in the endoplasmic reticulum. The H5 proteins were produced at relatively high concentrations in both systems. Following partial purification, haemagglutination and haemagglutination inhibition tests indicated that the conformation of the plant-produced HA variants was correct and the proteins were functional. The immunisation of chickens and mice with the candidate vaccines elicited HA-specific antibody responses. We managed, after synthesis of two versions of a single gene, to produce by transient and transgenic expression in plants, two variants of a highly pathogenic avian influenza virus HA protein which could have vaccine potential. This is a proof of principle of the potential of plant-produced influenza vaccines as a feasible pandemic response strategy for South Africa and other developing countries.

The spread of influenza A viruses is partially controlled and prevented by vaccination. The matrix protein 2 ectodomain (M2e) is the most conserved sequence in influenza A viruses, and is therefore a good potential target for a vaccine to protect against multiple virus subtypes. We explored the feasibility of a M2e-based universal influenza A vaccine candidate based on the highly pathogenic avian influenza A virus, H5N1. A synthetic M2e gene was human and plant codon optimised and fused in-frame with a sequence encoding the N-terminal proline-rich domain (Zera®) of the γ-zein protein of maize. Zera®M2e was expressed transiently in Nicotiana benthamiana and Sf21 baculovirus / insect cell expression systems, and Zera®M2e protein bodies (PBs) were successfully produced in both expression systems. The plant-produced Zera®M2e PBs were purified and injected into Balb/c mice. Western blot analysis using insect cell-produced Zera®M2e PBs and multiple tandem M2e sequences (5xM2e) fused with the avian influenza H5N1 transmembrane and cytosolic tail (5xM2e_tHA) confirmed the presence of M2e-specific antibodies in immunised mice sera. The immunogenicity of the Zera®M2e indicates that our plant-produced protein has potential as an inexpensive universal influenza A vaccine.

Human rotaviruses are the main cause of severe gastroenteritis in children and are responsible for over 500 000 deaths annually. There are two live rotavirus vaccines currently available, one based on human rotavirus serotype G1P[8], and the other a G1-G4 P[8] pentavalent vaccine. However, the recent emergence of the G9 and other novel rotavirus serotypes in Africa and Asia has prompted fears that current vaccines might not be fully effective against these new varieties. We report an effort to develop an affordable candidate rotavirus vaccine against the new emerging G9P[6] (RVA/Human-wt/ZAF/GR10924/1999/G9P[6]) strain. The vaccine is based on virus-like particles which are both highly immunogenic and safe. The vaccine candidate was produced in Nicotiana benthamiana by transient expression, as plants allow rapid production of antigens at lower costs, without the risk of contamination by animal pathogens. Western blot analysis of plant extracts confirmed the successful expression of two rotavirus capsid proteins, VP2 and VP6. These proteins assembled into VLPs resembling native rotavirus particles when analysed by transmission electron microscopy (TEM). Expression of the rotavirus glycoprotein VP7 and the spike protein VP4 was also tried. However, VP7 expression caused plant wilting during the course of the time trial and expression could never be detected for either protein. We therefore created three fusion proteins adding the antigenic part of VP4 (VP8*) to VP6 in an attempt to produce more appropriately immunogenic particles. Fusion protein expression in tobacco plants was detected by western blot using anti-VP6 and anti-VP4 antibodies, but no regular particles were observed by TEM, even when co-expressed with VP2. Our results suggest that the rotavirus proteins produced in N. benthamiana are candidates for a subunit vaccine specifically for the G9P[6] rotavirus strain. This could be more effective in developing countries, thereby possibly providing a higher overall efficacy for the existing vaccines. The production of rotavirus proteins in plants would probably result in lower manufacturing costs, making it more affordable for developing countries. Further investigation is required to evaluate the immunogenic potential of the VLPs and fusion proteins created in this study.

Rift Valley fever virus (RVFV), the causative agent of Rift Valley fever, is an enveloped single-stranded negative-sense RNA virus in the genus Phlebovirus, family Bunyaviridae. The virus is spread by infected mosquitoes and affects ruminants and humans, causing abortion storms in pregnant ruminants, high neonatal mortality in animals, and morbidity and occasional fatalities in humans. The disease is endemic in parts of Africa and the Arabian Peninsula, but is described as emerging due to the wide range of mosquitoes that could spread the disease into non-endemic regions. There are different tests for determining whether animals are infected with or have been exposed to RVFV. The most common serological test is antibody ELISA, which detects host immunoglobulins M or G produced specifically in response to infection with RVFV. The presence of antibodies to RVFV nucleocapsid protein (N-protein) is among the best indicators of RVFV exposure in animals. This work describes an investigation of the feasibility of producing a recombinant N-protein in Nicotiana benthamiana and using it in an ELISA. The human-codon optimised RVFV N-protein was successfully expressed in N. benthamiana via Agrobacterium-mediated infiltration of leaves. The recombinant protein was detected as monomers and dimers with maximum protein yields calculated to be 500–558 mg/kg of fresh plant leaves. The identity of the protein was confirmed by liquid chromatography-mass spectrometry (LC-MS) resulting in 87.35% coverage, with 264 unique peptides. Transmission electron microscopy revealed that the protein forms ring structures of ~ 10 nm in diameter. Preliminary data revealed that the protein could successfully differentiate between sera of RVFV-infected sheep and from sera of those not infected with the virus. To the best of our knowledge this is the first study demonstrating the successful production of RVFV N-protein as a diagnostic reagent by Agrobacterium-mediated transient heterologous expression in N. benthamiana. Preliminary testing of the antigen showed its ability to distinguish RVFV-positive animal sera from RVFV negative animal sera when used in an enzyme linked immunosorbent assay (ELISA). The cost-effective, scalable and simple production method has great potential for use in developing countries where rapid diagnosis of RVFV is necessary.