This technology could be used for rapid, flexible, scalable and cost-effective production of SARS-CoV-2 viral proteins, such as the full-length spike protein , spike protein fragments , or the nucleocapsid protein. These proteins, called antigens, bind specifically to SARS-CoV-2 antibodies in serological tests in Enzyme Linked Immunos orbent Assay or Lateral Flow Assay formats and can be used to test for the presence of these antibodies in human or animal samples. The soluble spike protein is a homotrimer comprised of 138 kDa highly glycosylated monomers, while the spike RBD is a 25 kDa fragment with only 2 glycosylation sites. For detection of the virus itself, proteins that bind to SARS-CoV-2 virus particles, such as the angiotensin converting enzyme-2 could also be produced in plants. Currently, these diagnostic reagents are being produced in mammalian cell culture , primarily using transient transfection processes due to the long times required to develop stable mammalian cell lines with high titers. However, the yields are quite low [∼5 mg/L for the full-length spike protein and >20 mg/L for the spike RBD ], and the processes require expensive media, transfection reagents, and expression enhancers, a source of mammalian cells with high viability, and have limited scalability. Depending on the host cell line and method used to introduce the DNA into the cells these processes may require more expensive facilities capable of handling human infectious agents. We propose that using plant tissue to produce proteins for COVID- 19 tests will be more cost-effective, faster, and scalable than traditional bio-manufacturing approaches. As an added bonus, this technology benefits agricultural producers by providing a market for crops that would otherwise be destroyed.In the transient agroinfiltration process,drying weed fresh plant biomass is submerged in a solution of genetically engineered bacteria, Agrobacterium tumefaciens , containing an expression cassette for production of the target protein.
In commercial scale facilities , whole plants grown indoors in trays under artificial lighting are used as the production host, however transient agroinfiltration and recombinant protein production in harvested plant biomass has also been demonstrated . For whole plants, the trays are turned upside down and the aerial portions of the plants are submerged in the agrobacterial solution while harvested plant biomass can be directly submerged. A moderate vacuum is applied to remove air from the interstitial spaces of the leaves for a few minutes, then the vacuum is released allowing the agrobacterial solution to infiltrate throughout the leaf tissue. The biomass is then incubated under controlled conditions, typically 5–7 days, to allow time for the agrobacteria to transfer the DNA production instructions to the plant host cells which make the protein using the plant host cell’s protein synthesis machinery. The plant biomass is then homogenized and the target protein is extracted, concentrated, and purified to the level required for the intended application. For diagnostic applications, an additional amino acid sequence can be included at the end of the protein to simplify purification using an affinity chromatography column. Although the production yields using this technology are protein dependent, they can reach up to 1 g/kg fresh weight biomass . We have previously presented a technoeconomic model for large scale production of monoclonal antibody therapeutic proteins produced in indoor-grown tobacco plants . This model is based on the Caliber Bio-therapeutics data generated during full-scale production for the DARPA “Blue Angel” project in the facility described by Holtz et al. . It can be used to provide a preliminary analysis of how a single facility like this could be used to produce SARS-CoV-2 diagnostic proteins in harvested lettuce.
Table 1 shows results of the calculations assuming the romaine lettuce feedstock is processed at the same throughput as N. benthamiana plant tissue for varying production levels in lettuce ranging from 1 mg purified antigen/kg FW to 100 mg purified antigen/kg FW after purification, assuming the same protein loss in downstream processing as for the mAb case. For the modeled facility with 1 day for vacuum infiltration , a 7 day incubation, and 1 day purification, the overall batch time from infiltration through bulk formulation is about 9 days. Higher throughputs could be achieved by increasing the number of vacuum infiltration units and downstream purification trains and staggering batches to infiltrate biomass more frequently than once a week . The cost of production of the antigen is assumed to be the same as the cost for processing one batch of N. benthamiana for mAb production [see Table 1 in Nandi et al. ] and adding $1.5M in feedstock costs to pay for the lettuce, coming to $28.4M/year not including depreciation, or $600K/batch. Note that these production costs also do not include any costs for transporting the harvested biomass from the farm to the processing facility. Agriculture is inherently a scalable system, especially with field grown plants that only require light, water and fertilizer, but particularly in situations where markets vanish rapidly in a large scale production infrastructure. Consider the fact that even with the lowest antigen production level it would only require about 36 acres of lettuce to produce antigens for 1 billion ELISA tests and about 120 acres for 1 billion LFA tests. Given published yields for full spike proteins it would require 60,000 L of mammalian cell culture to produce antigen for 1 billion ELISA tests and 200,000 L of mammalian cell culture to produce antigen for 1 billion LFA tests.
Although these scales are not unreasonably large, it should also be pointed out that there is also an urgent need for mammalian cell culture capacity for production of vaccines and therapeutics to treat COVID-19, in addition to the cost considerations described below. And while bioreactor working volumes up to 20,000 L are common for stable CHO suspension cultures, the largest scale that has been reported for transient transfection of mammalian cells is 100 L so antigen production for 1 billion ELISA tests would require 600 mammalian bioreactors runs; 20,000 bioreactor runs would be needed to produce antigen for 1 billion LFA tests. As antigen production yields increase for mammalian cell transient transfection processes, the required culture volumes and number of bioreactor runs will decrease proportionately. Ultimately, stable CHO lines with high antigen titers will be available but it will take time to develop these lines and scale up production processes. To meet current societal needs for diagnostic and therapeutic proteins, all bio-manufacturing platforms will be needed, including mammalian cell culture. Luckily, use of plant biomass has far better economies of scale and will be able to ramp up quickly to meet the 1 billion test target. Transient expression in plants is inherently fast, particularly when the plant biomass is already available, with an overall batch time, including purification, of about 9 days. To avoid transportation time and costs mobile vacuum infiltration units and prefabricated mobile clean room protein purification pods could be deployed onsite. Use of harvested plant biomass that is available saves a great deal of time compared with the time required to expand mammalian cells in culture for transfection and/or develop stable transgenic mammalian cell lines that express the antigen at high level. The overall batch time for transient transfection of mammalian cells at the 100 L bioreactor scale, not including downstream processing and purification, is about a month,vertical growing systems including the time required to expand cells prior to transfection . Cost is one of the main advantages of the proposed approach with the antigen production contributing cents or fractions thereof to the cost of the test. One cost driver is that plants can be grown outdoors while mammalian cells must be grown under aseptic conditions , as well as requiring expensive medium and transfection reagents. The cost estimates shown in Table 1 are an overestimate since they include costs associated with plant growth , as well as lighting during incubation and three chromatography steps which would not likely be needed for a diagnostic reagent. Waste biomass could be processed through anaerobic digestion to produce biogas energy to help power the facility as well as safe bio-fertilizer solids that could be returned to the farmer to offset raw materials costs. Assuming a mammalian cell culture media cost of $100/L, the cost of the medium alone, not including costs for labor, chromatography resins, buffers, utilities, etc., would contribute 0.6 cents per ELISA test and 2 cents per LFA at the stated expression level for the full spike protein of 5 mg/L. Our previous work has shown that annual production costs for monoclonal antibody production using the transient plant system are about half of the production costs for stable CHO mAb production facilities at similar production scales . We expect the cost advantage to be even more significant compared with mammalian cell culture transient transfection processes and use of lower cost field-grown plant biomass.Large multimeric glycoproteins are challenging to produce in any system and it is not clear, particularly for the full soluble spike trimer, whether expression levels of 100 mg/kg or greater can be obtained without significant process optimization.
As shown in Table 1, production yield plays a critical role in both antigen manufacturing costs and time to produce enough antigen for 1 billion tests. Another potential concern is feed stock variability associated with the use of field-grown plants which could introduce process challenges to ensure product quality and consistency. But field-grown plants have been used for medical applications for many years, for example the antimalarial drug artemisinin is still primarily obtained from field-grown Artemisia annua . In addition, field grown plants have also been used for transient production of therapeutics and vaccines. Over 25 years ago, the company Large Scale Biology, Inc. pioneered the use engineered plant viral vectors for transient production of therapeutic proteins and vaccines in field-grown tobacco, with extraction and purification of the bio-logics from the harvested plant tissue taking place in their indoor facility . In the scenario we propose for SARS-CoV-2 antigen production, transient agroinfiltration of harvested plant biomass would take place in an indoor facility, along with the extraction and purification of the antigen. While transient expression of recombinant proteins using agroinfiltration of harvested biomass has been demonstrated by our group and others at the lab scale in a variety of plants including N. benthamiana , sunflower , and lettuce , it has not yet been established at large scale. Additional work is needed to establish post-harvesting and post-infiltration handling and process conditions, as well as efficient expression systems for lettuce and other green leafy crops that are amenable to vacuum agroinfiltration. Equipment modifications and redesign will also be needed. For example, vacuum infiltrators will need to be redesigned/adapted to handle harvested biomass and process equipment and environmental controls for humidity and temperature during incubation will need to be established to preserve leaf health and protein production during the incubation phase. But, just as during World War II when large-scale stirred, aerated fungal fermenters were developed for the production of antibiotics by collaborative efforts of microbiologists and engineers , this could be the time to bring together agriculture and bio-manufacturing. It requires us to think creatively when it comes to how we make biologics, whether it is for industrial enzymes, biomaterials, diagnostics or therapeutics, perhaps broadening the plant-made proteins industry. This could also improve rural America’s trust in science , help keep farms in business and farm workers employed during the COVID-19 pandemic, and prepare us to respond to future outbreaks. Nebraska is located in the heart of the United States covering 76,878 square miles.The state has a population of 1.7 million people with 89% of the population being white and approximately 26% of Nebraskans are under the age of 18. For Nebraskans over the age of 18, 22.7% were current smokers as of 20022 , accounting for approximately 389,000 smokers. Since 1990, the prevalence rate of tobacco use has fluctuated from a high of 25.4% in 1990 to a low of 17.1% in 1992 and then reaching a plateau of about 22% through the late 1990s and early 2000s . Since 1995, the prevalence of adult tobacco use in Nebraska has remained about 1 percentage point below the national average. For decades, the tobacco industry has opposed any meaningful efforts to protect the health of Nebraskans through tobacco control.