Oct 6, 2010
In mid-2009, as a novel influenza virus swarmed across the planet with infections arising in more than seventy countries, the World Health Organization declared that a Phase 6[i] global pandemic was underway. Derived from swine, the H1N1 influenza spreads much like seasonal flu but has an antigenic profile largely unrecognized by existing antibodies. A year after the first H1N1 cases were reported, nearly three hundred pediatric deaths have been attributed to this particular strain of virus. Could some of these deaths have been prevented by vaccination? Current vaccine production methods take six months following initial isolation of the virus. Thanks to globalization and international travel, it took H1N1 just a few months to spread across nearly the entire world. Clearly, if met with a truly lethal virus, existing methods are not nimble enough to respond in time.
The current standard for vaccine production is the chicken egg model. Eggs make a useful viral production powerhouse because they have a large, self-contained allantoic compartment (the egg white) in which the virus can replicate. Every flu season, strains of influenza are selected by the WHO’s Global Influenza Surveillance Network as the most likely infection candidates for the coming season[ii] and then used to infect millions of chicken eggs. Over the course of a six-month incubation period, the virus is allowed to replicate within the egg. Following incubation, the virus is purified from the cell material and inactivated by one of a variety of different methods usually involving a fixative like formaldehyde and a detergent to disrupt the viral membrane. From the viral debris, key antigenic surface proteins such as hemagglutinin and neuraminidase are isolated and formulated with fillers and preservatives like gelatin, sucrose, and thimerosol (a mercury derivative)[iii]. The proteins are what the body’s immune system recognizes when an infection occurs normally. This final formulation is then reconstituted prior to administration by injection or inhalation. One or two eggs’ worth of virus are needed to constitute a single vaccinating dose. Thus, to vaccinate the entire candidate population, ~400 million eggs would have to be viably processed. This is limited not only by the enormity of the scale, but also by the fact that the eggs can spoil during the six-month incubation. At present, vaccine producers combine to generate approximately half of the consumer demand. In response to the recent H1N1 pandemic, five companies (CSL Limited, ID Biomedical Corporation of Quebec, Novartis Vaccines and Diagnostics Limited, Sanofi Pastuer Inc., and MedImmune LLC) were approved by the FDA to produce vaccines against the new influenza strain[iv]. All of these manufacturers used the chicken egg method.
Understanding the above description of conventional vaccine manufacturing, we see that the active components are merely the isotype-specific coat proteins hemagglutinin and neuraminidase. With modern methods of molecular biology and protein synthesis, it is possible to generate these very proteins in their isotype-specific formats in a variety of systems. Such is the goal of three different new companies focused on alternative methods of vaccine production. Neugenesis, Protein Sciences Corp., and Fraunhofer USA each have their own unique approach to antigen production that could change both the source and speed of vaccine production.
Neugenesis[v], based in Burlingame, CA, is using filamentous fungi to make the necessary proteins. Filamentous fungi grow as multinuclear spindles called hyphae and in a variety of sexually differentiated forms. Neugenesis uses heterokaryotic fungi transfected with DNA encoding antigenic proteins to produce the vaccine material. With thorough international patent protection[vi], they are poised to be a major contributor to vaccine production in the future. Neugenesis is banking on their NeuBIOSTM system and its ability to generate antigens at levels vastly exceeding conventional methods, allowing for production of greater numbers of vaccines to meet global demand. NeuBIOS can apparently produce hemagglutinin antigen at 100 mg/L, equivalent to 10 million doses in ~1,500 L. They also suggest that they can go from cell-banked virus to full production in just seven weeks.
Protein Sciences Corp.[vii] in Meriden, CT is using an equally unique approach to vaccine production. Like Neugenesis, they aim to produce recombinant influenza coat proteins to serve as vaccine material. However, they have a quite different platform; they use cells from spodoptera frugiperda, a type of caterpillar with a baculovirus expression system. Baculovirus commonly infects insect cells. In this system, DNA encoding hemagglutinin antigen is combined with DNA from baculovirus and used to infect the caterpillar cells. In this particular embodiment, the hemagglutinin is associated with the insect cell walls, which helps in the purification protocol entailing membrane disruption and a series of chromatography steps. Their patented process[viii] allows them to isolate 99% pure hemagglutinin with yields of ~20 mg/L[ix] and formulate a vaccine that contains no egg components (which sometimes leads to allergic reactions) nor any preservatives such as the aforementioned thimerosol (eliminates mercury component). Further, they have demonstrated the ability to deglycosylate antigens (remove post-translational sugar modifications on the protein that are species-specific), which may prove useful for reasons discussed below. FluBlok®, a seasonal flu vaccine, is currently under FDA review, and the company has two other vaccines currently in Phase II clinical trials.
Fraunhofer USA[x], or more specifically their Center for Molecular Biotechnology located in Newark, DE, is growing plants to make vaccines. They suggest that just about any plant can be used[xi], but it seems one of their favorites is Nicotiana benthamiana, a tobacco relative. They are motivated to antigen production by viral infection, similar to the process used by Protein Sciences Corp., with a plant-specific virus like tobacco mosaic virus. The desired antigen is genetically encoded into plant virus DNA or RNA, depending on the type of plant virus, and then plant virus is manually rubbed onto the plant. This initiates a local infection that can then spread to the rest of the plant. As the infection spreads, each newly infected cell begins synthesizing more of the encoded antigen. Maintenance of the influenza antigen gene in the plant viral genome is achieved by splitting the genes required for infection onto separate plasmids, one of which includes the target gene. The antigen may be expressed as a fusion with various plant viral components such as coat proteins or as a separate protein with independent localization signal[xii]. One interesting benefit of this expression system is that for certain proteins, perhaps even vaccines, it may be desirable to formulate without purification from the plant host. In other words, you could be eating a salad and be getting your vaccine at the same time.
One area of potential concern for this particular approach is glycosylation. Glycosylation refers to the post-translational process that proteins undergo in eukaryotic cells in which certain residues are modified with sugar groups. Different eukaryotic organisms (humans, yeast, fungi) have different patterns of glycosylation that are often recognized by the immune system. It might be that immune recognition of fungal glycosylation might induce an even more effective response, but since viruses use their human host cells to replicate, a neutralizing antibody specific to a fungal glycosylation site might be ineffective. This issue is likely to be a common concern for most alternative eukaryotic production systems.
At this point, none of the above vaccines have been approved by the FDA or other regulatory agencies. Most approved biotherapeutics are manufactured in mammalian cells, so this may represent a barrier to entry for these vaccines. Assuming concerns such as alternative glycosylation patterns can be overcome, alternative vaccine production methods would likely become the predominant approach for seasonal vaccines due to economic considerations and for spontaneous, unforeseen pandemics by necessity. These methods offer antigen yields that should also prove sufficient to support the massive, global immunizations necessary to stem the spread of a new virus.
[i] World Health Organization. Current WHO phase of pandemic alert. (2009) http://www.who.int/csr/disease/avian_influenza/phase/en/
[ii] World Health Organization. Recommended composition of influenze virus vaccines for use in the 2010 southern hemisphere influenza season. (September 2009) http://www.who.int/csr/disease/influenza/200909_Recommendation.pdf
[iii] Sanofi Pasteur. 449/454 Influenza A (H1N1) 2009 Monovalent Vaccine. (10 September 2009) http://www.fda.gov/downloads/BiologicsBloodVaccines/Vaccines/ApprovedProducts/UCM182404.pdf
[v] Neugenesis. www.neugenesis.com
[vii] Protein Sciences Corporation. www.proteinsciences.com
[viii] Smith GE, Volvovitz F, Wilkinson BE, et al. Vaccine comprising a baculovirus produced influenza hemagglutinin protein fused to a second protein. United States Patent and Trademark Office (January 1999) Patent #5858368
[x] Fraunhofer USA Center for Molecular Biotechnology. www.fraunhofer-cmb.org