The most straightforward approach to vaccine development would be based on inactivated or attenuated strains of SARSCoV-2, but the production of sufficient material generally takes longer than is the case for subunit vaccines, high-level containment would be necessary to grow the virus before attenuation/inactivation, and the candidates would carry a risk of reacquired virulence . For subunit vaccines, target antigens must be selected carefully. Research on the original SARS-CoV strain indicated that the N protein is highly conserved among coronavirus families, including strains responsible for mild respiratory tract infections, thus suggesting the possibility of developing a universal vaccine. However, antibodies induced by N proteins did not provide protective immunity; likewise, the M and E proteins elicited only weak protective responses . These studies helped to confirm the S protein as the most suitable target for eliciting a neutralizing humoral response.The entry of coronaviruses into host cells is facilitated by the S protein, which assembles into homotrimers on the virus surface . The S protein comprises two functional subunits: S1, which binds to the host cell receptor, and S2,drainage collection pot which facilitates the fusion of the viral and host cell membranes.
For many coronaviruses, the S protein is cleaved at the boundary between the S1 and S2 subunits and mostly remains non-covalently bound in the pre-fusion conformation . Hence, the uptake of coronaviruses into host cells is a complex process that requires receptor binding and proteolytic processing of the S protein to stimulate membrane fusion and viral uptake .Companies currently developing COVID-19 vaccines are mainly expressing variants of the SARS-CoV-2 S1 protein or RBD. The S1 proteins of SARS-CoV and SARS-CoV-2 are heavily glycosylated, with an approximately equal mixture of complex and high-mannose glycans . It is unclear whether plant-type complex glycans would affect the efficacy of a recombinant SARS-CoV-2 S-protein vaccine expressed in plants. High-mannose glycans are generally conserved across higher eukaryotes, so it could be expected that at least some high-mannose glycans will be added during the expression of the antigen in plants. Furthermore, it is not clear whether sialic acid plays a role in host-receptor interactions. This is not generally present on native or recombinant plant glycoproteins, although engineered plant varieties that produce sialylated proteins have been described . Virus-like particles displaying SARS-CoV-2 antigens are larger than subunit vaccines, promoting recognition and internalization by antigen-presenting cells and thus triggering an adaptive immune response. Furthermore, the regular array of epitopes acts as pathogen-associated molecular patterns to induce strong cellular and humoral responses .
VLPs are readily produced at scale in plants by molecular farming . The Medicago VLP platform is a prime example and has previously been used to produce millions of doses of seasonal influenza vaccines . Furthermore, iBio is also using a proprietary system to develop VLP-based vaccines in N. benthamiana plants.Given the time required to develop and test a COVID-19 vaccine, the possibility that a vaccine may not be effective in all populations due to the variability of immune responses, and the likelihood that SARS-CoV-2 will mutate, we foresee an ongoing demand for therapeutic proteins, such as mAbs, immunoadhesins, interferons, and antivirals, to either target the virus itself or reduce the severity of the associated acute respiratory syndrome .Several recombinant mAbs and antibody cocktails against COVID-19 are currently undergoing clinical development for therapeutic and prophylactic applications, including REGNCoV-2 , CSL312 , LY-CoV555 , and TYO27 . Many of the mAbs in development target the S-protein, aiming to block interactions with its receptor, angiotensin-converting enzyme 2 . Efforts to exploit convalescent sera from patients who recovered from COVID-19 have helped identify antibodies with neutralizing potential. For example, Eli Lilly/ AbCellera identified such an antibody in a blood sample from one of the first United States patients who recovered from the disease. The mAb was developed into LY-CoV555, a potent, neutralizing IgG1 that binds the S protein. In collaboration with NIAID, the product began Phase III clinical evaluation in high-risk assisted living facilities in August 2020.21 Most COVID-19 antibody products in development are produced in mammalian cells, but antibodies were among the first products of molecular farming in plants and many different mAb products have been expressed, including complex secretory IgA .
The dose of a mAb or mAb cocktail needed for the prevention or treatment of COVID-19 is currently unclear. About 9 g of the ZMapp cocktail was needed per treatment against Ebola virus and in a subsequent clinical study , but that dose level was selected from the outcome of studies in non-human primates , which enabled rapid deployment under the compassionate use protocol and did not benefit from dose optimization studies in humans. Assuming similar doses, manufacturing scalability is likely to be a key challenge in the production of COVID-19 antibodies. The scaling up of conventional bioreactors is particularly challenging due to changes in mixing, mass transfer, and heat exchange, whereas transient expression in plants can be scaled in a linear manner because each plant is effectively an independent bioreactor, equating to a process of numbering up by increasing the plant inventory and throughput of the facility. Similarly, cost will be an important consideration. In 2013, total sales of mAbs produced in mammalian cell bioreactors amounted to ~€48.5 billion for 8,182 kg of product, with an average sales price of ~€5,957 g−1 . Production costs and capital expenses for the transient expression of mAbs in plants are estimated to be at least 50% lower than mammalian cell culture production facilities , allowing manufacturers to reduce sales prices while still making some profit or providing these therapeutics at cost, and saving taxpayer resources.Another promising therapeutic approach is the use of plants to produce immunoadhesins . Such molecules combine the virus-binding region of a receptor, in this case ACE2, with the immunoglobulin Fc domain . The ACE2 component acts as a decoy to bind SARS-CoV-2 via the S protein, preventing it from engaging with native ACE2 on the surface of human cells, while the Fc region confers a longer circulatory half-life and provides effector functions that promote viral clearance,square plastic pot as well as facilitating product purification by Protein A affinity chromatography during manufacturing. Immunoadhesins form dimers via disulfide linkages between Fc domains, increasing their avidity when binding the S protein. One advantage of this strategy is that if the coronavirus mutates to escape binding to the immunoadhesins, it would similarly lose affinity for native ACE2, reducing its infectivity. Likewise, the SARS virus that re-emerged in 2003–2004 had a lower affinity for ACE2 than the original isolate, resulting in less severe infections and no secondary transmission . An additional advantage of this strategy is that exogenous ACE2 would compensate for lower ACE2 levels in the lungs during infection, thereby contributing to the treatment of acute respiratory distress. Several companies in the United States and the EU have developed recombinant ACE2 and ACE2-Fc fusion proteins for preclinical and clinical testing, although all these products are currently produced in mammalian cell lines . The impact of plant-specific complex glycans on the ability of ACE2-Fc to bind the RBD has been studied using molecular dynamic simulations and illustrates the important role that glycosylation may play in the interaction between the S protein and ACE2 .Griffithsin is a lectin that binds high-mannose glycans, and is currently undergoing clinical development as an antiviral against HIV-1.
However, it also binds many other viruses that are pathogenic in humans, including HSV , HCV , Nipah virus , Ebola virus, and coronaviruses including SARS-CoV and MERS , and as recently determined, also SARSCoV-2. A clinical product in development by University of Louisville is currently manufactured in N. benthamiana by Kentucky Bio-processing using a TMV vector. The API is also undergoing preclinical development as a nasal spray for use as a non-vaccine prophylactic against coronaviruses, with clinical evaluation planned for 2020 . This candidate PMP antiviral could be deployed under the EUA pathway if found effective in controlled clinical studies. Griffithsin is an interesting example of a product that is ideally matched to plant-based manufacturing because it is naturally produced by a marine alga. Griffithsin has been expressed with limited success in E. coli and tobacco chloroplasts, but better results have been achieved by transient expression in N. benthamiana using A. tumefaciens infiltration or TMV vectors, with expression levels of up to 1 g kg−1 fresh mass and recoveries of up to 90% . A TEA model of griffithsin manufactured in plants at initial commercial launch volumes for use in HIV microbicides revealed that process was readily scalable and could provide the needed market volumes of the lectin within an acceptable range of costs, even for cost-constrained markets . The manufacturing process was also assessed for environmental, health, and safety impact and found to have a highly favorable environmental output index with negligible risks to health and safety.In addition to COVID-19 PCR tests, which detect the presence of SARS-CoV-2 RNA, there is a critical need for protein-based diagnostic reagents that test for the presence of viral proteins and thus report a current infection, as well as serological testing for SARS-CoV-2 antibodies that would indicate prior exposure, recovery, and possibly protection from subsequent infection. The design and quality of the binding reagents , along with other test conditions such as sample quality, play a key role in establishing the test specificity and selectivity, which determine the proportion of false positive and false negative results. Although the recombinant protein mass needed for diagnostic testing is relatively small , the number of tests needed for the global population is massive, given that many individuals will need multiple and/or frequent tests. For example, 8 billion tests would require a total of ~2.5 kg purified recombinant protein, which is not an insurmountable target. However, although the production of soluble trimeric full-length S protein by transient transfection in HEK293 cells has been improved by process optimization, current titers are only ~5 mg L−1 after 92 h . Given a theoretical recovery of 50% during purification, a fermentation volume of 1,000 m3 would be required to meet the demand for 2.5 kg of this product. Furthermore, to our knowledge, the transient transfection of mammalian cells has only been scaled up to ~0.1 m3 . The transient expression of such protein-based diagnostic reagents in plants could increase productivity while offering lower costs and more flexibility to meet fluctuating demands or the need for variant products. Furthermore, diagnostic reagents can include purification tags with no safety restrictions, and quality criteria are less stringent compared to an injectable vaccine or therapeutic. Several companies have risen to the challenge of producing such reagents in plants, including Diamante , Leaf Expression Systems , and a collaborative venture between Plant Form, Cape Bio Pharms, Inno-3B, and Microbix.Resilience is the state of preparedness of a system, defining its ability to withstand unexpected, disastrous events , and to preserve critical functionality while responding quickly so that normal functionality can be restored . The concept was popularized by the 2011 Fukushima nuclear accident but received little attention in the pharmaceutical sector until COVID-19. Of the 277 publications retrieved from the National Library of Medicine22 on July 9th 2020 using the search terms “resilience” and “pandemic,” 82 were evenly distributed between 2002 and 2019 and 195 were published between January and July 2020. Resilience can be analyzed by defining up to five stages of a resilient system under stress, namely prevent, prepare, protect, respond, and recover . Here, prevent includes all measures to avoid the problem all together. In the context of COVID-19, this may have involved the banning of bush meat from markets in densely populated areas . The prepare stage summarizes activities that build capacities to protect a system and pre-empt a disruptive event. In a pandemic scenario, this can include stockpiling personal protective equipment but also ensuring the availability of rapid-response bio-pharmaceutical manufacturing capacity. The protect and respond stages involve measures that limit the loss of system functionality and minimize the time until it starts to recover, respectively. In terms of a disease outbreak, the former can consist of quarantining infected persons, especially in the healthcare sector, to avoid super-spreaders and maintain healthcare system operability . The response measures may include passive strategies such as the adjustment of legislation, including social distancing and public testing regimes, or active steps such as the development of vaccines and therapeutics .