How Does Science Work?

Risk as Perception

Risk as Science

The Risk Assessment Paradigm

How Foods from Biotech Crops are Evaluated for Human Safety

How Biotech Crops are Evaluated for Environmental Safety in the United States

The Benefits and Risks of Producing Pharmaceutical Proteins in Plants

Originally Printed in: Risk Management Matters 2(4):28-33.
Reprinted with permission, copyright Risk Group 2004.

Leslie M. Shama and Robert K. D. Peterson
Agricultural and Biological Risk Assessment,
Montana State University, Bozeman, MT 59717

Protein-based pharmaceuticals traditionally used for the treatment of disease have been made through the expression of protein in bacterial, fungal, and mammalian cell cultures.  Recently, the possibility to produce more diverse and complex pharmaceutical proteins in plants has reached the laboratory benches of scientists and companies worldwide.  These pharmaceuticals are known as plant-made or plant-based pharmaceuticals.  In this article, we will discuss pharmaceutical proteins, how genetic engineering makes it possible to produce proteins in plants, why plants are desirable for the production of pharmaceutical proteins, and which plant systems are being considered for their production.  Additionally, we will discuss risk and regulatory issues associated with this new technology.

Proteins as Pharmaceuticals

Proteins are essential to all living organisms for function, structure, and regulation of the body.  Protein development in a cell begins with DNA transcribing into RNA, and RNA translating into proteins. Proteins are made up of amino acids that are arranged in different combinations and lengths.  The differences in arrangements and lengths of amino acids determine the function of the protein.  Some examples of proteins include hormones, enzymes, and antibodies.

Many people suffer from infectious, inflammatory, and cardiovascular diseases — and these numbers are growing.  Protein-based drugs are the fastest growing class of drugs for the treatment of these diseases in humans and other diseases in animals.  The reasons for this are because the numbers of people with diseases such as diabetes are growing and new technologies are making proteins easier to produce.  The current methods of production of proteins for pharmaceutical application (mammalian, bacterial, and fungal cell cultures) are predicted to fall short of demand in the near future (Rogers 2003). 

Insulin was the first pharmaceutical protein produced using genetically engineered bacteria (Thomas et al. 2002).  Insulin originally was isolated from cows and pigs that were slaughtered for food.  This method was inefficient and caused some patients to develop allergies from the animal-derived insulin.  Today it is made from the human gene that codes for the insulin protein and is expressed and cloned in the bacterium, Escherichia coli.  Large quantities of E. coli are now grown in fermentation vats to make tons of human insulin available to the growing number of diabetic patients.

Plant Transformation

Genetic engineering also has made it possible to use plants as factories for pharmaceutical protein production.  Plant-made pharmaceuticals are made by inserting a segment of DNA that encodes the protein of choice into plant cells.  The plants or plant cells are essentially factories used to produce the desired proteins and are only grown for the purpose of pharmaceutical applications.

There are two common methods of transformation (the process by which DNA from one organism is incorporated into the DNA of another organism) that have been established through biotechnology to produce transgenic plants, which, in turn, could be used to create the plants used to make pharmaceutical proteins.  The transformation techniques include the Agrobacterium tumefaciens-mediated transformation system and biolistics, also called particle bombardment.

Agrobacterium tumefaciens is a bacterium that naturally infects plants and causes crown gall disease.  It is very useful for the production of transgenic plants because it has the amazing ability to transfer a segment of its DNA, called T-DNA, into the nucleus of the plant cells.  The T-DNA from A. tumefaciens is then integrated into the plant and transcribed, causing crown gall disease.  Scientists have used A. tumefaciens to their advantage by inserting their DNA of interest between the T-DNA to create a plant with qualities such as herbicide resistance, or pesticide resistance, as well as many others (Nester et al. 1984, Binns and Thomashow 1988). Crown gall disease does not occur in the plant after being transformed by A. tumefaciens because the plasmid has been constructed to disarm the tumor-inducing properties that cause the disease, and instead it expresses the DNA of interest.

Biolistics, or the Particle Bombardment Transformation System, employs the use of metal particles either of gold or tungsten.  The metal particles are coated with the DNA that is to be transferred into the plant cells.  Using a pressurized system, a plastic bullet coated with metal and DNA particles is released from the biolistics machine when the pressure is released.  The bullet is shot, and within seconds, stopped by a shield causing the metal coated DNA particles to be knocked off the bullet. The particles are then ultimately forced to be inserted into the plant cells below.

Once the genes have been transferred into the cells, the cells expressing the gene of interest must be selected.  In most cases, a selectable marker is included with the DNA construct, which will confer resistance to an antibiotic or herbicide.  The plant cells that survive the antibiotic or herbicide application are the ones that have successfully received the gene of interest (Purdue Agricultural Biotechnology 2004). 

Plant Systems vs. Mammalian and Bacterial Cell Cultures

Plants and plant cells offer several advantages in the production of proteins for pharmaceutical purposes.  Unlike mammalian cells and bacterial cells, plants are capable of post-translational modification and other assembly steps that are needed for biological activity in complex multi-component proteins such as antibodies.  This in turn allows for more diversity and possibilities of drugs that could be conceived from the proteins produced in plants (Rogers 2003, Goldstein and Thomas 2004).  When a protein is glycosylated, this means a certain sugar molecule is added to its structure, causing it to fold in a specific and unique orientation.  The way the protein is folded determines its function and activity.  Plants are capable of producing more complex proteins with no assistance because they are capable of glycosylation on their own, whereas mammalian cells need assistance to glycosylate complex proteins.  The system needed to maintain and create complex proteins in mammalian cells is expensive and difficult to maintain, which makes plant systems a more attractive source for protein production.  (Fischer et al. 2003). 

Mammalian cells are difficult and expensive to grow; therefore, bacterial cells are used to produce proteins when glycosylation is not required.  Bacterial cells still have the potential of being contaminated, and require facilities to be built to keep up with the growing demand of proteins needed for drug production.  This is why plants are an excellent alternative to mammalian and bacterial cell cultures; they have to ability to glycosylate complex proteins and are inexpensive to grow in large quantities. 

Plant Options

Plant-made pharmaceutical production has the potential to provide large amounts of protein for the pharmaceutical industry.  The types of plants used for pharmaceutical protein production depend largely on their final application.  To date, the plants that have successfully been transformed include tobacco, potato, tomato, corn, soybeans, alfalfa, rice, and wheat.  However, it is important to emphasize that no commercially available pharmaceutical products are currently produced in plants (Ma et al. 2003). 

Tobacco was the first plant to be genetically engineered and has the advantage of being used as a plant biopharmaceutical for exactly that reason; the methods for gene transfer and expression are well established for this plant.  Tobacco can also be cropped multiple times per year, giving it an edge over other plants in producing a large biomass (Fischer et al. 2003).  The protein of interest can also be targeted to the seed.   In tobacco, up to one million seeds can be made on a single plant (Daniell et al. 2001).  One concern associated with using tobacco as a protein factory is that it has toxic alkaloids associated with it, which could be present in the final drug that is administered to the patient.  Another concern is that if the protein is targeted for the leaf tissue, it must be frozen or dried and then processed immediately, making it difficult to harvest and store for preservation of the protein (Ma et al. 2003). 

Legumes, such as alfalfa and soybean, and cereal crops, such as corn and rice, have been considered as ideal candidates for plant biopharmaceuticals because the protein can be targeted to accumulate in the seed and the seed can be harvested and stored for an extended amount of time.  Legumes are good candidates because they naturally produce large amounts of protein in their seeds, making the final protein recovery much larger.  When considering a cereal crop as a protein factory, yield is the main quality of importance, as well as ease of transformation and speed of production scale-up (Ma et al. 2003).

Protein Application

The application of the proteins made in plants include, antibodies, vaccines, hormones, enzymes, interleukins, interferons, and human serum albumins.  Monoclonal antibodies have the potential to be among the first protein pharmaceuticals commercially produced in plants (Rogers 2003).  Antibodies are produced in vertebrate immune systems to recognize and bind to antigens with amazing specificity.  Because antibodies possess this specificity they can be used as diagnostic tools as well as prevention and treatment of diseases (Fischer et al. 2003).  There are at least six types of plant-derived recombinant antibodies that have moved to the preclinical stage of testing, and one more advanced antibody in phase II clinical trial (Ma et al. 2003).  A possible use for monoclonal antibodies is to one day attach a chemotherapeutic agent onto the antibody and allow it to specifically bind to a tumor cell and eventually the chemotherapeutic agent will destroy the tumor cell (Rogers 2003).   

Another application for plant-made pharmaceuticals is the production of vaccines.  Subunit vaccines are made up of specific macromolecules that induce a protective immune response against a pathogen.  Subunit vaccine technology has increased the safety of administering vaccines because it does not involve the use of live or weakened viruses. Currently, it is very expensive to produce subunit vaccines as well as to store them because they are not heat-stable.  Because the vaccines are not heat-stable, this limits where they can be sent for use and unfortunately makes them unavailable in the developing countries where vaccines are needed the most.  Producing the vaccines in plants eliminates the heat-stability issue because the vaccinogenic plant tissue can be administered raw, dried, or in an encapsulated form; all of these forms can be stored and shipped at room temperature (Sala et al. 2003).  The risk of contamination with animal pathogens during production is also eliminated (Walmsley et al. 2000).  Oral delivery of the vaccine is another reason that vaccines produced in plants are an attractive possibility because they can eliminate injection-related risks.  Overall, the fact that the vaccines can be stored as seeds are advantageous because large amounts of vaccines can be produced in limited time and storage is less of an issue because the seed is a stable form that will not degrade the protein over time.

The choice of the crop will determine the way the vaccine is finally administered because only some plants can be consumed raw whereas others must be processed.  With processing there is the potential that heat or pressure treatments could destroy the protein.  Cereal crops are an attractive species for expressing subunit vaccines because they can produce proteins in their seeds, which are stable for long storage periods.  For animal vaccines the plant choice could be chosen based on what is eaten as a major part of its diet, therefore eliminating the need to process the protein and the risk of destroying it as well (Daniell et al.  2001).




In the U.S., a three-agency coordinated framework including the Food and Drug Administration (FDA), the United States Department of Agriculture (USDA), and the Environmental Protection Agency (EPA) is responsible for regulating all products made from biotechnology. Currently, the FDA regulates pharmaceuticals in the U.S., but with the introduction of plant-made pharmaceuticals the USDA plays a key role.  Each agency has its own regulatory scope. For example, the FDA is responsible for overseeing the production of pharmaceuticals for their safety to humans and animals.  The USDA evaluates genetically engineered plants and animals to assess whether they will pose unacceptable risks to U.S. agriculture and the environment.  For plant-based pharmaceuticals, EPA will play a lesser role, but will help ensure safety to the environment. The activities of the FDA and USDA include not only products that are registered, but also experimental products grown in laboratory, field, and greenhouse environments. Specific regulations are currently evolving, but at this time the specific responsibilities and activities of the two agencies remain unclear (Peterson and Arntzen 2004).

The production of proteins in plants to be used in the pharmaceutical industry will largely follow the same regulatory requirements that have already been established for non-plant produced pharmaceuticals. However, there are key challenges for the regulation of plant-based pharmaceuticals. These challenges primarily are associated with the fact that some production systems will include manufacturing pharmaceutical proteins in the open environment. Further, many production systems will involve food. These are unique aspects of pharmaceutical manufacturing –– aspects that will present novel risks that must be considered. As Peterson and Arntzen (2004) stated, “Because of this, questions of intra- and inter-species gene flow, allergen exposure to the public and non-target organism exposure come into play.”

Potential Risks

The potential risks associated with plant-based pharmaceuticals include: pollen transfer to related species, contamination of non-transgenic crops intended for the consumption by humans, allergic reactions to the drugs produced from the genetically engineered plant, and persistence of genetically engineered material to persist in the environment and accumulate in non-target organisms (Daniell et al. 2001).  Risk assessment of plant-made pharmaceuticals should be reviewed on a case-by-case basis because the plants used to produce proteins each have different risks associated with them (Peterson and Arntzen 2004). 

To address the out-crossing potential of transgenic crops to a receptive crop species or weedy relative, physical isolation distances have been considered.  This can mean containment in a greenhouse setting or buffer zones between fields of the same species.  Some species have enclosed flower structures, reducing the distance needed in between fields and therefore are being evaluated as good candidates for producing plant-made pharmaceuticals.  In the case of corn –– a wind pollinated crop –– because it has many benefits for being a plant biopharmaceutical, the tassels can be manually removed and male-sterile varieties could be used to contain the pollen.  The transgenic plants can also be planted at different times from food crops to make sure they flower at different times, which decreases the potential for pollen transfer (Thomas et al. 2002).

According to Goldstein and Thomas (2004), “virtually every pharmaceutical product currently on the market can cause allergic reactions in some people.”  The FDA approved the drugs because their benefits outweighed their allergenic potential.  This will be the same approach taken on plant-made pharmaceuticals; the FDA will be concerned if, and only if, the plant-made pharmaceuticals are more allergenic than pharmaceuticals not derived from plants (Rogers 2003).  However, there is a question of the allergic risk for people not taking the drug but happen to inhale or topically receive pollen from the genetically engineered plant.  Will the contact from the genetically engineered plant cause a reaction or immunity?  Because these proteins are in the environment does not necessarily make them environmental contaminants or even human health hazards.  The specificity of the proteins produced in plants potentially makes the risk negligible.

Non-target organisms are organisms not intended to be affected by the plant-produced protein. These organisms include insects, mammals, birds, fish, plants, and other wildlife in and around the area where the genetically engineered plant is growing.  Animals potentially could be affected by genetically engineered plants, by eating or landing on the plant or plants nearby that have the pollen from the genetically engineered plants on them.  There are ways to prevent or reduce the risk of these non-target organisms from being affected by the genetically engineered plant.  Planting border crops that are not genetically engineered near the engineered crops, or containing the plants in greenhouse facilities are a couple preventative ways to keep non-target organisms unharmed.  Fortunately, many of the plant-made pharmaceutical proteins currently produced or being considered are species-specific (such as antibodies) which reduces their effects (Goldstein and Thomas 2004). 

Risk Assessment

Uncertainties in the evolving regulatory environment and the need to evaluate the human-health and environmental risks of plant-based pharmaceuticals on a case-by-case basis necessitate using a robust, transparent science-based methodology. Therefore, Peterson and Arntzen (2004) recommended using the established paradigm of environmental risk assessment. Additionally, Wolt and Peterson (2000) argued that the risk assessment framework was sufficiently robust to evaluate risks from most applications of biotechnology.

Environmental risk assessment is a formalized basis for the objective evaluation of risk in a manner in which assumptions and uncertainties are clearly considered and presented. Risk assessment proceeds in a systematic, stepwise fashion that includes the following five steps: (1) problem formulation; (2) hazard identification; (3) dose-response relationships; (4) exposure assessment; and (5) risk characterization. Hazard and dose are considered in relation to exposure to determine risk or what additional information is needed to calculate or refine risk estimates (NRC 1983).


Plants used as protein factories to assist the pharmaceutical industry in drug manufacturing have the advantage of producing proteins which could not be made using other systems, reducing the cost of production, and increasing the amount of proteins available to be used to make drugs.  With this increased availability and potentially lower cost, more patients will be able to receive the drugs they need.  Plant systems are capable of producing large quantities of protein and have an advantage over mammalian and bacterial cell cultures because of the ease of scale-up that is possible with the equipment and land already in existence for these plants. These systems should be assessed for risk using the risk assessment paradigm. Risk analysis (which includes risk assessment, risk management, and risk communication) is a flexible and robust framework, making it ideal for addressing the myriad issues associated with this new technology.


References Cited

Binns, A.N., Thomashow, M.F.  1998.  Cell biology of Agrobacterium infection and transformation of plants.  Annual Review of Microbiology 42:575-606.

Daniell, H., Streatfield, S. J., Wycoff, K.  2001.  Medical molecular farming: production of antibodies, biopharmaceuticals and edible vaccines in plants.  Trends in Plant Science 6:219-226

Fischer, R., Twyman, R. M., Schillberg, S. 2003.  Production of antibodies in plants and their use for global health. Vaccine 21:820-825.

Goldstein, D.A., Thomas, J.A.  2004.  Biopharmaceuticals derived from genetically modified plants. QJM: An International Journal of Medicine 97:705-716.

Ma, J.K-C., Drake, P.M.W., Christou, P.  2003.  The production of recombinant pharmaceutical proteins in plants.  Nature Reviews Genetics 4: 794-805.

Nester, E.W., Gordon, M.D., Amasino, R.M., Yanofsky, M.F.  1984.  Crown gall: a molecular and physiological analysis. Annual Review of Plant Pathology 35:387-413.

 [NRC] National Research Council. 1983. Risk Assessment in the Federal Government: Managing the Process.  Washington, DC: National Academy Press.

Peterson, R.K.D., Arntzen, C.J.  2004.  On risk and plant-based biopharmaceuticals.  Trends in Biotechnology 22:64-66.

Purdue Agricultural Biotechnology. 2004.  Moving Genes into Plants.

Rogers, K.K.  2003.  The Potential of Plant-Made Pharmaceuticals.

Sala, F., Rigano, M. M., Barbante, A., Basso, B., Walmsley, A.M., Castiglione, S.  2003.  Vaccine antigen production in transgenic plants: strategies, gene constructs and perspectives. Vaccine 21:803-808.

Streatfield, S.J., Lane, J.R., Brooks, C.A., Barker, D.K., Poage, M.L., Mayor, J.M., Lanphear, B.J., Drees, C.F., Jilka, J.M., Hood, E.E., Howard, J.A.  2003. Corn as a production system for human and animal vaccines. Vaccine 21:812-815.

Thomas, B.R. Van Deynze, A., Bradford, K.J.  2002.  Production of Therapeutic Proteins in Plants. Agricultural Biotechnology in California Series 8078. University of California-Davis. 12pp.

Walmsley, A.M., Arntzen, C.J.  2000.  Plants for delivery of edible vaccines. Current Opinion in Biotechnology 11:126-129.

Wolt, J.D., Peterson, R.K.D.  2000.  Agricultural biotechnology and societal decision-making: the role of risk analysis. AgBioForm 3:39-46.

Comments? © 2001-2005