Field of Protein Dreams
Researchers have long sought to use plant-based bioreactors to produce biopharmaceuticals. Recent advances mean that dream might soon come true.
Discovery and Development, James Netterwald, PhD Senior Editor
Plants have always been useful to the pharmaceutical industry but have never had the power that they do now that the era of plant-based bioreactors, the brainchild of modern biotechnology, has been ushered in. "Some of the earliest ideas of using plants to produce novel proteins came from plant virologists,” says Roger Beachy, PhD, president of The Donald Danforth Plant Center, St. Louis. Those earlier scientists looked at the possibility of using plant viruses to carry new genetic information into a host plant, the result of which was temporary production of novel proteins. This type of transient expression of a foreign protein represented the beginning of plant biotechnology.
The first company to capitalize on this new technology was Large-Scale Biology Corp., a Vacacille, California-based company that developed a number of technologies that allowed for the production of large amounts of foreign protein from the white tobacco plant in the mid-1980s. Although the company profited from the transient expression technology for about 10 to 12 years, it encountered marketing problems that caused them to finally close their doors in December, 2005.
The next wave in plant biotechnology came in the form of the transgenic plant, which quickly demonstrated its advantages over the transient system. Transgenic plants allow for stable, continuous expression of novel proteins because the protein itself is encoded in a genetically-heritable transgene. “Once you have the transgenic plant, you plant it in a greenhouse, a cave, a growth chamber, or in a field, and it will just produce the transgenic protein,” says Beachy.
Another advantage is that the cost of producing a transgenic protein in a transgenic plant is substantially lower than producing the same protein in a fermentation-based bioreactor which may use Escherichia coli, yeast, or mammalian cell culture. “Most of these fermenters require a lot of up-front cost, sometimes up to $100 million,” says Henry Daniell, PhD, a professor at the University of Central Florida and technical founder of Chlorogen Inc., St. Louis. The major cost-reducing factor is that the transgenic plants can be grown on inexpensive land or in low-cost facilities. “Some estimate that the incremental cost savings are as low as 10% and others estimate them to be as high as 50%,” says Beachy.
The safety of products derived from fermenter-based product is also an issue. “The fermenters have to be maintained under sterile conditions. Every year we hear about contamination of them by the flu virus and things like that.” says Daniell. This is definitely a concern, especially for mammalian cell bioreactors, because they can become infected with human and animal pathogens that can ultimately contaminate the protein product. This is not a concern for plant bioreactors, however, so they are able to produce proteins with a high safety profile.
Overcoming production limits
The concept of plant bioreactors emerged from the fact that plants have a very high biomass and, therefore, have the potential to produce substantially higher yields of recombinant protein compared with other expression systems. However, this is not the case with transgenic plants. In fact, compared with the 10% to 30% yields of the plant virus vector system, the transgenic protein makes up only 0.1% to 1% of the total protein produced by the plant. This low yield is due to the fact that transgenic plants are prepared by injecting the transgene into plant cell nuclei, a method that allows for only one copy of this gene per cell.
The ability to increase the biomass of the bioreactor is an important factor in choosing a bioreactor. “I think the most obvious advantage of a plant bioreactor is the ability to scale-up very quickly,” says Karen Oishi, PhD, chief scientist and vice president of technology development at Chlorogen. The low cost of creating a plant bioreactor makes it very easy to scale-up production because manufacturers can increase biomass by simply growing more plants. Examples of plants that can produce a very high biomass include tobacco, rice, corn, and soybean.
The solution to this yield problem has come from an obvious place—inside the plant itself. “We insert our genes into the chloroplast organelle of the tobacco plant, where there could be as many as 200 chloroplasts per cell,” says Oishi.
Each chloroplast contains approximately 50 copies of the plant’s extrachromosomal DNA, which is maternally inherited. So there is the potential to stably transform every chloroplast in the cell, and, therefore, every cell in the plant, with the transgene. Consequently, when the transgene integrates into chloroplast DNA, the chloroplast becomes the transgenic protein factory of the cell.
Of course, there are some basic rules to follow when constructing the transgene. For example, if the transgene is being expressed in corn, then the transgene should be under the control of a corn-specific promoter that targets the protein expression to a specific organ
(e.g., seed) in the plant.
Biological limitations
There are biological limitations when expressing the transgenic protein in plant systems. “You need to address things like codon bias, leader sequences, and other translation-related issues,” says Beachy. Product stability is also an issue for expressing most transgenic proteins in plants. Luckily, the chloroplast environment is similar to that of E. coli, so when trying to express transgenic proteins in chloroplasts, protein stability can be first tested in E. coli. Protein stability can also be achieved by expressing the transgenic proteins between the plant cell membrane and the cell wall, an area which has a low concentration of proteases and other hydrolytic enzymes.
Another concern germane to plant bioreactors is that of posttranslational modifications such as glycosylation. “Plants do glycosylate transgenic proteins, but they tend to put extra sugars in them,” says Oishi. Overall, the process and cellular machinery for glycosylation of proteins is conserved in plants, but not 100% conserved. Glycosylation of specific amino acids in the protein’s structure is absolutely essential for some animal proteins to function. So, if an animal protein that requires specific glycosylation for bioactivity is produced in a transgenic plant that lacks the ability to properly glycosylate that protein, then that protein may be inactive.
“One of the problems faced by chloroplasts is that they do not glycosylate proteins, so this system has a limitation of not being able to produce glycoproteins,” says Oishi. But the chloroplast system has the ability to produce very complex proteins, multimers, peptides, and very large proteins, among others, the capability which has been demonstrated by Chlorogen and many academic laboratories. Bioactivity is not always an issue for human proteins made in chloroplasts. Daniell has made many human blood proteins, including interferon, insulin, and growth factors, in chloroplasts and has not found any differences in bioactivity.
“There are efforts ongoing to put into plants the enzymes necessary to give animal-like modifications of animal proteins,” says Beachy. This requires genetic engineering of the host plant to give it the ability to make such posttranslational modifications and then reengineering of the plant to insert the transgene.
In most cases, the transgenic protein must be purified after it is produced. There are many components in plants, such as phenolics and alkaloids, which can interfere with purification of the transgenic protein. For example, phenolics can cause the protein to precipitate or alter the protein’s charge, such that it behaves differently during purification chromatography.
Of course, the concern for most manufacturers is whether this kink in the purification step will impact the cost of the manufacturing process. “The purification costs are not substantially different than if you were to purify from a CHO cell line in a cell culture system,” says Beachy. “I think the issues of production are being overcome.”
Prove yourself
To date, no plant-produced, transgenic proteins have been approved by the US Food and Drug Administration (FDA) for human use. However, there are several candidates that are currently in clinical trials, so it is unclear how these products will fare in the regulatory process. “I think that the FDA has certain regulations about not allowing the transgenics to get into people’s food chain,” says Chenming Zhang, PhD, assistant professor of biological systems engineering at Virginia Polytechnic Institute and State University, Blacksburg. So the transgenic plants must be grown in an isolated area away from any food crops. Tobacco is an example of a non-food, non-feed crop used as a plant bioreactor.
“Tobacco is a wonderful system. . . . It produces a lot of biomass and produces a lot of seeds for replanting so that scale-up can be done easily.” Also, because it is a non-food, non-feed crop, there is less of a chance for the transgene to escape the transgenic plant and enter the human food chain. This is also not an issue for the chloroplast system because the transgene DNA in chloroplasts is extrachromosomal, and therefore there is no chance for it to be present in pollen and consequently no chance for it to outcross the food crop.
“The main thing is that people want to know what they are eating,” says Zhang. This does not mean that food crops, such as corn, rice, and soybean, cannot be used as plant bioreactors. They are already being used as experimental bioreactors, and as long as their growth is isolated to prevent contamination of the food chain, their use will likely continue.
Although some fear of consuming human proteins produced by plants should be expected, in general, people in the US consume a great deal of genetically modified foods. “In fact, the US is a world leader in this, with 97% of our soybean product being genetically-modified,” says Daniell. However, what if people could eat a vaccine instead of administering it by injection? “If you grow the vaccine in E. coli or in yeast, you would have to inject the vaccine. However, by using edible plants, such as banana, tomato, lettuce, and others, people can get access to the vaccine simply by eating the plants,” says Zhang. This type of vaccine delivery would be ideal for impoverished countries, where vaccination is a luxury, not a necessity.
Vaccines for bio weapons
There are several vaccines currently in development, most of which are directed toward biological weapons, such as anthrax, cholera, and plague. Vaccines against biological weapons will be used to protect military personnel and civilians from succumbing to a bioterrorist attack. “We have published that one acre of transgenic plants can produce 360 million doses of anthrax vaccine,” says Daniell, who adds that this amount would be enough to protect the entire US population.
The plant-produced vaccine comes in response to the shortage of fermenter-produced anthrax vaccine. Also, military personnel refuse to take the current vaccine for fear of being poisoned by the Bacillus anthracis toxin, a frequent contaminant of this vaccine. The plant-produced vaccine is free of this toxin, so it is more likely to be readily accepted by military personnel.
The biggest challenge will be getting this vaccine through the regulatory process and approved by the FDA. “All of the current FDA guidelines are written for the fermentation technology . . . so there is really no guideline for the number of units per acre and so forth,” says Daniell. Consequently, new guidelines will need to be written and new approval processes put in place to facilitate the approval of products made in plant bioreactors.
In collaboration with Daniell, the Defense Advanced Research Projects Agency (DARPA), an arm of the United States Department of Defense, has agreed to make millions of doses of the plant-produced anthrax vaccine in one of its cGMP facilities. DARPA has also agreed to walk this new anthrax vaccine through the regulatory process.
Although the plant bioreactor industry has many challenges ahead, these challenges do not differ from those faced by their predecessors. “I think it will be a viable system; it is just a matter of time before somebody gets it done,” says Zhang. Once the technical problems are resolved and one of these proteins is approved by the FDA, then the promise of the plant bioreactor will come to pass.
Discovery and Development, James Netterwald, PhD Senior Editor
Plants have always been useful to the pharmaceutical industry but have never had the power that they do now that the era of plant-based bioreactors, the brainchild of modern biotechnology, has been ushered in. "Some of the earliest ideas of using plants to produce novel proteins came from plant virologists,” says Roger Beachy, PhD, president of The Donald Danforth Plant Center, St. Louis. Those earlier scientists looked at the possibility of using plant viruses to carry new genetic information into a host plant, the result of which was temporary production of novel proteins. This type of transient expression of a foreign protein represented the beginning of plant biotechnology.
The first company to capitalize on this new technology was Large-Scale Biology Corp., a Vacacille, California-based company that developed a number of technologies that allowed for the production of large amounts of foreign protein from the white tobacco plant in the mid-1980s. Although the company profited from the transient expression technology for about 10 to 12 years, it encountered marketing problems that caused them to finally close their doors in December, 2005.
The next wave in plant biotechnology came in the form of the transgenic plant, which quickly demonstrated its advantages over the transient system. Transgenic plants allow for stable, continuous expression of novel proteins because the protein itself is encoded in a genetically-heritable transgene. “Once you have the transgenic plant, you plant it in a greenhouse, a cave, a growth chamber, or in a field, and it will just produce the transgenic protein,” says Beachy.
Another advantage is that the cost of producing a transgenic protein in a transgenic plant is substantially lower than producing the same protein in a fermentation-based bioreactor which may use Escherichia coli, yeast, or mammalian cell culture. “Most of these fermenters require a lot of up-front cost, sometimes up to $100 million,” says Henry Daniell, PhD, a professor at the University of Central Florida and technical founder of Chlorogen Inc., St. Louis. The major cost-reducing factor is that the transgenic plants can be grown on inexpensive land or in low-cost facilities. “Some estimate that the incremental cost savings are as low as 10% and others estimate them to be as high as 50%,” says Beachy.
The safety of products derived from fermenter-based product is also an issue. “The fermenters have to be maintained under sterile conditions. Every year we hear about contamination of them by the flu virus and things like that.” says Daniell. This is definitely a concern, especially for mammalian cell bioreactors, because they can become infected with human and animal pathogens that can ultimately contaminate the protein product. This is not a concern for plant bioreactors, however, so they are able to produce proteins with a high safety profile.
Overcoming production limits
The concept of plant bioreactors emerged from the fact that plants have a very high biomass and, therefore, have the potential to produce substantially higher yields of recombinant protein compared with other expression systems. However, this is not the case with transgenic plants. In fact, compared with the 10% to 30% yields of the plant virus vector system, the transgenic protein makes up only 0.1% to 1% of the total protein produced by the plant. This low yield is due to the fact that transgenic plants are prepared by injecting the transgene into plant cell nuclei, a method that allows for only one copy of this gene per cell.
The ability to increase the biomass of the bioreactor is an important factor in choosing a bioreactor. “I think the most obvious advantage of a plant bioreactor is the ability to scale-up very quickly,” says Karen Oishi, PhD, chief scientist and vice president of technology development at Chlorogen. The low cost of creating a plant bioreactor makes it very easy to scale-up production because manufacturers can increase biomass by simply growing more plants. Examples of plants that can produce a very high biomass include tobacco, rice, corn, and soybean.
The solution to this yield problem has come from an obvious place—inside the plant itself. “We insert our genes into the chloroplast organelle of the tobacco plant, where there could be as many as 200 chloroplasts per cell,” says Oishi.
Each chloroplast contains approximately 50 copies of the plant’s extrachromosomal DNA, which is maternally inherited. So there is the potential to stably transform every chloroplast in the cell, and, therefore, every cell in the plant, with the transgene. Consequently, when the transgene integrates into chloroplast DNA, the chloroplast becomes the transgenic protein factory of the cell.
Of course, there are some basic rules to follow when constructing the transgene. For example, if the transgene is being expressed in corn, then the transgene should be under the control of a corn-specific promoter that targets the protein expression to a specific organ
(e.g., seed) in the plant.
Biological limitations
There are biological limitations when expressing the transgenic protein in plant systems. “You need to address things like codon bias, leader sequences, and other translation-related issues,” says Beachy. Product stability is also an issue for expressing most transgenic proteins in plants. Luckily, the chloroplast environment is similar to that of E. coli, so when trying to express transgenic proteins in chloroplasts, protein stability can be first tested in E. coli. Protein stability can also be achieved by expressing the transgenic proteins between the plant cell membrane and the cell wall, an area which has a low concentration of proteases and other hydrolytic enzymes.
Another concern germane to plant bioreactors is that of posttranslational modifications such as glycosylation. “Plants do glycosylate transgenic proteins, but they tend to put extra sugars in them,” says Oishi. Overall, the process and cellular machinery for glycosylation of proteins is conserved in plants, but not 100% conserved. Glycosylation of specific amino acids in the protein’s structure is absolutely essential for some animal proteins to function. So, if an animal protein that requires specific glycosylation for bioactivity is produced in a transgenic plant that lacks the ability to properly glycosylate that protein, then that protein may be inactive.
“One of the problems faced by chloroplasts is that they do not glycosylate proteins, so this system has a limitation of not being able to produce glycoproteins,” says Oishi. But the chloroplast system has the ability to produce very complex proteins, multimers, peptides, and very large proteins, among others, the capability which has been demonstrated by Chlorogen and many academic laboratories. Bioactivity is not always an issue for human proteins made in chloroplasts. Daniell has made many human blood proteins, including interferon, insulin, and growth factors, in chloroplasts and has not found any differences in bioactivity.
“There are efforts ongoing to put into plants the enzymes necessary to give animal-like modifications of animal proteins,” says Beachy. This requires genetic engineering of the host plant to give it the ability to make such posttranslational modifications and then reengineering of the plant to insert the transgene.
In most cases, the transgenic protein must be purified after it is produced. There are many components in plants, such as phenolics and alkaloids, which can interfere with purification of the transgenic protein. For example, phenolics can cause the protein to precipitate or alter the protein’s charge, such that it behaves differently during purification chromatography.
Of course, the concern for most manufacturers is whether this kink in the purification step will impact the cost of the manufacturing process. “The purification costs are not substantially different than if you were to purify from a CHO cell line in a cell culture system,” says Beachy. “I think the issues of production are being overcome.”
Prove yourself
To date, no plant-produced, transgenic proteins have been approved by the US Food and Drug Administration (FDA) for human use. However, there are several candidates that are currently in clinical trials, so it is unclear how these products will fare in the regulatory process. “I think that the FDA has certain regulations about not allowing the transgenics to get into people’s food chain,” says Chenming Zhang, PhD, assistant professor of biological systems engineering at Virginia Polytechnic Institute and State University, Blacksburg. So the transgenic plants must be grown in an isolated area away from any food crops. Tobacco is an example of a non-food, non-feed crop used as a plant bioreactor.
“Tobacco is a wonderful system. . . . It produces a lot of biomass and produces a lot of seeds for replanting so that scale-up can be done easily.” Also, because it is a non-food, non-feed crop, there is less of a chance for the transgene to escape the transgenic plant and enter the human food chain. This is also not an issue for the chloroplast system because the transgene DNA in chloroplasts is extrachromosomal, and therefore there is no chance for it to be present in pollen and consequently no chance for it to outcross the food crop.
“The main thing is that people want to know what they are eating,” says Zhang. This does not mean that food crops, such as corn, rice, and soybean, cannot be used as plant bioreactors. They are already being used as experimental bioreactors, and as long as their growth is isolated to prevent contamination of the food chain, their use will likely continue.
Although some fear of consuming human proteins produced by plants should be expected, in general, people in the US consume a great deal of genetically modified foods. “In fact, the US is a world leader in this, with 97% of our soybean product being genetically-modified,” says Daniell. However, what if people could eat a vaccine instead of administering it by injection? “If you grow the vaccine in E. coli or in yeast, you would have to inject the vaccine. However, by using edible plants, such as banana, tomato, lettuce, and others, people can get access to the vaccine simply by eating the plants,” says Zhang. This type of vaccine delivery would be ideal for impoverished countries, where vaccination is a luxury, not a necessity.
Vaccines for bio weapons
There are several vaccines currently in development, most of which are directed toward biological weapons, such as anthrax, cholera, and plague. Vaccines against biological weapons will be used to protect military personnel and civilians from succumbing to a bioterrorist attack. “We have published that one acre of transgenic plants can produce 360 million doses of anthrax vaccine,” says Daniell, who adds that this amount would be enough to protect the entire US population.
The plant-produced vaccine comes in response to the shortage of fermenter-produced anthrax vaccine. Also, military personnel refuse to take the current vaccine for fear of being poisoned by the Bacillus anthracis toxin, a frequent contaminant of this vaccine. The plant-produced vaccine is free of this toxin, so it is more likely to be readily accepted by military personnel.
The biggest challenge will be getting this vaccine through the regulatory process and approved by the FDA. “All of the current FDA guidelines are written for the fermentation technology . . . so there is really no guideline for the number of units per acre and so forth,” says Daniell. Consequently, new guidelines will need to be written and new approval processes put in place to facilitate the approval of products made in plant bioreactors.
In collaboration with Daniell, the Defense Advanced Research Projects Agency (DARPA), an arm of the United States Department of Defense, has agreed to make millions of doses of the plant-produced anthrax vaccine in one of its cGMP facilities. DARPA has also agreed to walk this new anthrax vaccine through the regulatory process.
Although the plant bioreactor industry has many challenges ahead, these challenges do not differ from those faced by their predecessors. “I think it will be a viable system; it is just a matter of time before somebody gets it done,” says Zhang. Once the technical problems are resolved and one of these proteins is approved by the FDA, then the promise of the plant bioreactor will come to pass.
