Published online 31 May 2007
Published in Crop Sci 47:1255-1262 (2007)
© 2007 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
GENOMICS, MOLECULAR GENETICS & BIOTECHNOLOGY
Methods for Growing Nonfood Products in Transgenic Plants
John A. Howarda,* and
Elizabeth E. Hoodb
a Applied Biotechnology Institute, Bldg. 36, Cal Poly State Univ., San Luis Obispo, CA 93407
b Arkansas State Univ., P.O. Box 2760, State Univ., AR 72467
* Corresponding author (jhoward{at}appliedbiotech.org).
 |
ABSTRACT
|
|---|
The relatively high cost of producing select industrial and pharmaceutical products in traditional hosts has led many groups to investigate the benefits of plant-based production systems. Transgenic plants can offer significant cost advantages for select products, but the advantage can be eroded occasionally by the cost of confinement to segregate them from commodity crops. The conceptual similarities and differences between transgenic plant production systems and other transgenic hosts are discussed in regard to regulations and public perception. A system of regulated transgenic production is described that is based on dedicating an area solely to industrial products. This system would take advantage of surrounding crops that are also grown for industrial applications to offset the cost of growing the regulated transgenic plants. Examples are given that demonstrate the economic advantages of the system while concurrently creating a clearer distinction between food and nonfood products. This system may also increase public confidence, as it is modeled on current systems used to produce regulated transgenic products in nonplant hosts.
Abbreviations: FDA, Food and Drug Administration • GRAS, generally regarded as safe.
|
INTRODUCTION
|
|---|
WITH THE ADVENT of recombinant technology, microbes quickly became the hosts of choice to produce transgenic proteins for industrial and pharmaceutical applications. Methods for producing proteins in plants were developed later but were first used on commodity crops to confer traits for agronomic improvement. More recently, transgenic crops are being considered for their ability to produce proteins used in industrial and pharmaceutical applications (Kusnadi et al., 1997; Fischer et al., 1999; Hood and Howard, 1999; Fischer and Emans, 2000; Daniell et al., 2001; Giddings, 2001; Hood, 2002; Hood and Howard, 2002a; Ma et al., 2003; Howard and Hood, 2005). For centuries plants have been a source for industrial and pharmaceutical products, and the use of transgenic plants can be taken as the next logical step in the evolution of a more efficient production system. The hope is that this new generation of industrial and pharmaceutical products from plants can enable cleaner environmental solutions and more affordable medicine. This hope has spurred the formation of several groups to examine the use of plants as factories for therapeutics, oral vaccines, and industrial enzymes and fill the void when other production platforms are too costly or inadequate for expressing particular proteins.
Although these new uses of plants are well intended, they have raised public concern that transgenic nonfood products can inadvertently mix with the food supply. This comes at a time when the public has already become sensitized to the perceived safety concerns regarding transgenic plants used for trait improvement in food crops. To ensure safety and gain public confidence, the USDA has adopted stringent guidelines (USDA, 2003) distinct from those used for crops intended to enter the food supply. Compliance with these guidelines can be quite burdensome and costly to the growers of these new products. Nevertheless, companies have encouraged regulation because of the need to have the public accept this new technology and because the cost can often be absorbed by the relatively high price for some of these products.
While the associated costs for current practices of containment may be acceptable for high-priced pharmaceuticals, they can significantly limit the production of many industrial products. Industrial products have relatively lower price ceilings, and the additional cost of production based on regulations cannot always be absorbed. This has led us to evaluate ways to maintain the current safety practices of the USDA guidelines, address the public perception of safety, and be cost-effective.
|
COMPARISON OF TRANSGENIC PROTEIN PRODUCTION SYSTEMS
|
|---|
Inherent Safety of the Compound
The first concern is for product safety and to evaluate the inherent toxicity of the molecule if a person receives direct exposure through intended uses. Science-based models exist for predicting the potential toxic effects based on dosage and are currently used to evaluate pharmaceutical or food compounds by the Food and Drug Administration (FDA) in the USA. These evaluations provide a baseline safety assessment of the compound to determine the inherent toxicity of the molecule. This aspect of safety is the same for all production systems.
Final Product Safety
The second aspect of safety includes any unintended compounds that may be introduced into the final product during the purification or production process. This aspect is host specific and also depends on the level of purification in the final product. Some products are highly purified and contain only trace amounts of their host proteins, while other products may be only partially purified and contain a large amount of their host proteins. In addition, certain animal hosts can carry disease, thus testing is warranted to ensure selected animal viruses or toxins are absent in the final product.
It can be argued for hosts that are already in the food chain, that they have a distinct advantage in that they already have generally regarded as safe (GRAS) status. GRAS status can apply to a number of transgenic protein production systems; for example food organisms such as baker's yeast may cause much less concern than other fungi not in the food chain. Moreover, some fungi are known to produce toxins, making them even more problematic. This same logic also applies to plants in that commodity food crops may have an inherent advantage for safety as opposed to crops not in the food chain or known to produce toxins. Therefore, in relation to final product safety, little conceptual difference is apparent between plants and other production systems, but there can be a great deal of difference as to the specific organism used in either case.
Unintended Exposure in Nonproduct Uses
Production practices must ensure that the transgenic product be confined and that it does not appear in unintended places. Regulations are in place for all types of production systems to restrict the movement of a transgenic protein, including proper labeling and transport of the transgenic protein, as well as the confinement of the host to limit its ability to reproduce and generate transgenic protein. There are differences as to how to achieve this confinement depending on the system, but all methods include genetic and physical barriers to limit the host's reproduction outside of the production site.
Unintended exposure due to production practices has caused the most concern for plant-based production because of the potential for the transgenic proteins to end up in our food or feed supply. However, the possibility of unwanted compounds in our food supply is not new with transgenic protein production. Many products such as insecticides, herbicides, and packaging material can end up in our food supply, along with mycotoxins, insects, or other animal matter. Some of these unintended compounds can be highly toxic, and while it is the goal to keep these out altogether, there is a realization that minute quantities may enter the food supply despite our best efforts. Therefore, safety models have led to regulations based on tolerances or action levels to indicate that below certain levels there is no cause for concern if extraneous materials remain (USEPA, 1989, USEPA, 1992).
What is different for plant-based transgenic protein production is that these existing models have not been used to set limits of possible exposure, and the assumption has been that any amount of transgenic protein material creates a problem, regardless of whether it is safe or not, and regardless of whether it can be detected. There is reason to believe that the current models used for other safety assessments can be adapted for transgenic plants producing nonfood compounds. One model has been proposed specifically to evaluate potential safety concerns of unintended exposures when producing pharmaceutical products in plants (Howard and Donnelly, 2004). Other models and refinements are likely to evolve over time as more products emerge. Some of these potential products are unlikely to show a human health risk from inadvertent exposure since they do not constitute a novel compound entering the food supply, primarily because they are already present in the food we eat. It is possible that some of these potential products may eventually be deregulated, but even so, these would still require confinement until a thorough safety assessment has been conducted. Some potential plant-produced compounds under development are based on the sequence of animal-derived proteins also found in the food chain, and yet others may be completely novel. We advocate that all potential products undergo safety assessments, and in no way are we suggesting that any system of production be used to circumvent any part of the current guidelines.
Public Perception
Any successful production system must account for the public's perceived concerns as well as science-based safety concerns. The most frequently cited issue from the public with regard to transgenic plant products is that these could be intermixed with our food supply and pose a significant hazard. Even when safety evaluations can demonstrate that a product poses no significant risk, it is not in itself enough to gain public acceptance. To address the public's perceived safety concern, we compared the differences between using transgenic plants versus other organisms that currently are used in transgenic protein production and are readily accepted by the public.
One concept that frequently emerges is the use of food organisms as hosts for nonfood products, raising the public's concern. This concept, however, is in direct conflict with current production practices that use yeast and eggs (food sources) to produce products such as industrial enzymes, vaccines, and pharmaceuticals. These nonplant food hosts have been used successfully for decades. The food host argument diverts attention from the real issue, which is to ensure that transgenic nonfood products remain outside the food system, rather than which host is used for production. Growing industrial or pharmaceutical products in nonfood hosts does not eliminate the possibility for crops or microbes to intermix with the food supply. Therefore, using only a nonfood host will be unlikely to solve the public's concern and may raise new issues for product safety.
One difference between plant-based production systems and other hosts is the site of production. Microbes producing nonfood products are usually housed in a building separate from those microbes producing food ingredients. This helps provide physical containment and gives the public confidence that the food supply is segregated from industrial organisms. In practice, however, it is not the building itself containing the microbes that provides security, since microbes can become airborne and escape even more easily than plants. It is the many other practices that are used inside the building that are enforced by regulatory agencies that keep the microbes contained. Perhaps even more importantly from the public's perception is that there are no food products being produced at the same site as the nonfood products.
One perceived concern for plants has been that they cannot be confined since they are not grown inside a building; therefore, the public is at more risk. This concept has led to some groups investigating contained facilities such as greenhouses, plant cell cultures, or aquacultures, as well as underground caves. The advantage of providing containment inside a building more closely resembles what is done for nonplant hosts. However, growing plants inside a building adds substantial cost and can be limiting for products requiring large-scale production and/or very low costs, thus eroding one of the reasons for using plants instead of nonplant hosts.
In practice, confinement can also be accomplished in the field by a variety of methods, many of which are suggested in the USDA guidelines. These include genetic, temporal, and geographic practices. Physical separation can be obtained by limiting the growth of certain crops within a prescribed distance of the regulated transgenic crops, as well as the use of border rows. However, one major conceptual difference between the systems used for field-grown plants and those used for microbes is that in the case of microbes there is a dedicated facility that is not readily interchanged for making food and industrial products. In the case of field-grown plants, the USDA guidelines are written to protect the food supply, assuming that in any given season or location, plants can be grown for food, feed, or industrial applications. The practice of using the same land to grow regulated products one year and commodity crops in subsequent years requires costly steps for producers to adhere to the guidelines. This also fuels the public perception that these products can easily be intermixed with food crops.
An Alternative Model for Growing Transgenic Protein in Plants
As an alternative to having crop land flexible to grow all types of products, land can be dedicated solely for the production of transgenic plants for industrial products. In this case, at a dedicated location all of the equipment, personnel, and practices employed would be solely for growing transgenic plants for industrial applications every year. This model is similar in concept to what has been done for microbial and cell culture systems in that the location is not used interchangeably for the production of food and nonfood transgenic products. The dedicated area required for the transgenic product may be achieved by one grower with a large parcel of land, or multiple growers may collectively work together as a cooperative (co-op) and impose this restriction.
Increasing safety margins and gaining public confidence are very worthy goals, but what makes this approach particularly attractive is that it can also increase the economic outcome. We have attempted to illustrate these advantages using specific examples. We have selected corn (Zea mays L.) to represent a crop producing a transgenic nonfood product. Corn is used today for industrial, food, and feed applications and is also being developed to produce bioindustrials and biopharmaceutical transgenic products. There are three variations of this concept that are illustrated below for corn, but many of the conclusions can be equally applied to other crops as well.
|
CASE 1: DEDICATING ACREAGE TO GROW ONLY THE REGULATED TRANSGENIC CROP FOR A BIOPHARMACEUTICAL OR BIOINDUSTRIAL PRODUCT
|
|---|
The first example is that a grower simply dedicates the land year after year to transgenic production for a nonfood use. This simple and obvious case has the advantages that a dedicated growing area could- create a greater distinction for the public between food and nonfood transgenic products, including separate locations, specialized growers, separate dedicated handling equipment, and easy identification;
- eliminate the problem of disposal of volunteer plants that develop the following year after planting a transgenic industrial crop (volunteer plants from scattered seeds currently need to be destroyed to eliminate the potential to intermix with food crops that can be later planted on the same acreage);
- reduce fears that seed dropped in the handling and transport could be easily intermixed with seed entering the food chain, since no food crops will be planted on the land;
- provide local growers with specialized expertise and a way to increase their revenues; and
- be more cost-effective for growing the nonfood crops.
With regard to economic incentives, the biggest cost savings may be that the land used for growing the transgenic nonfood product does not need to remain fallow the following year. Therefore, the revenue that would have been lost due to the regulations that prohibit the planting of a commodity crop can now be captured by planting another industrial crop. Assuming a grain yield of 150 bushels/acre for commodity corn at $2.00/bushel, this model provides added revenue of $300/acre. In addition, the cost associated with monitoring and destroying volunteers would be reduced. These economic incentives may not be significant when producing high-value pharmaceutical products, which can have values in the thousands of dollars per acre for the raw material, but may be essential when producing much lower unit cost industrial products.
While this approach may seem obvious and has some very appealing features, it is not necessarily attractive to growers. In order for this to work at the grower level, there must be a constant demand for the transgenic industrial crop every year, otherwise it is no different than the present case, and growers must switch to nonregulated crops in years where there is no need for industrial production. Therefore, while this scenario is very feasible for the future, it has limited practical applications immediately unless growers are guaranteed significant production of the specialized crops for years in advance.
|
CASE 2: PRODUCTION OF A TRANSGENIC PHARMACEUTICAL PRODUCT WITHIN A DEDICATED ETHANOL-PRODUCING AREA
|
|---|
The second example employs the increasing trend of using commodity corn for industrial applications. In particular, industrial applications have grown as a percentage of total corn acres in the last decade largely due to the increase in the production of grain ethanol. Production of ethanol from grain was over 4.3 billion gallons per year in 2005 (http://www.ksgrains.com/ethanol/useth.html; accessed 29 Nov. 2005; verified 11 Mar. 2007) and is expected to increase again in 2006 to account for 15% of grain produced in the US. Growth of this crop for industrial applications is nationwide and intermixed with growth for food and feed. Because transportation cost is a concern, most of the ethanol facilities are located close to grain production (http://www.iowabeefcenter.org/content/USAMapEthanolPlantsJuly05.pdf; accessed 29 Nov. 2005; verified 11 Mar. 2007), but this does not restrict the same crops from being used for food and feed. Our proposed model differs from what is currently practiced because growing locations surrounding an ethanol facility would be dedicated solely for industrial use, and the grain could not be used in food applications. Owners of the grain ethanol facility would be required to contract with the surrounding growers. This may not be as difficult as it initially sounds since some of the grower cooperatives are also owners of the local ethanol facilities (http://www.iowacorn.org/ethanol/ethanol_8.html; accessed 29 Nov. 2005; verified 11 Mar. 2007), making this level of coordination relatively straightforward for select areas. Since most nonfood transgenic products require a very small amount of land for production compared to what is grown for food and feed, only a select few grain ethanol facilities would need to utilize this production model compared to the vast majority of grain ethanol plants that would not see any change in their practice.
To illustrate how this proposed system may work, we have selected a representative pharmaceutical product, aprotinin, which can be produced in corn grain. Although the human safety risks of inadvertent exposure may be insignificant for aprotinin, as it is already in the food chain (present in beef), it will still require strict adherence to the USDA guidelines to gain public confidence without deregulation of its host. We have modeled a production scheme for aprotinin in the proposed dedicated growing system (Fig. 1). The key assumptions and calculations are shown in Table 1. Production of the transgenic aprotinin crop on relatively small acreage represents one specific application of the crop, while the major industrial application for the corn crop is grain ethanol that requires most of the surrounding acreage and can be supplied using commodity corn.
In this example, the pharmaceutical produced in corn accounts for 0.001% of the current corn crop in the United States. Therefore, growing of this regulated transgenic crop will make no appreciable change in the scenario for growing corn in general, including having no effect on the corn supply or the price of commodity corn. Rather, this specialized crop pertains only to a small select group of growers. This model can provide direct savings for growing the crop as shown in Case 1 by using the land to plant this same crop the following year rather than allowing it to remain fallow. This scenario also offers additional features. Specifically, the growers can increase their acreage substantially without penalty if there is an increase in demand for aprotinin, or decrease the production of the regulated crop when demand is low and instead plant commodity corn that would go to the production of ethanol. This practice also provides the added benefit of having a large buffer zone for the regulated transgenic crop with enough acreage to justify the dedicated equipment that would be required for regulated transgenic crops. Therefore, this is more practical for growers' needs than Case 1.
|
CASE 3: PRODUCTION OF THE INDUSTRIAL ENZYME CELLULASE USED FOR STOVER PRODUCTION OF ETHANOL
|
|---|
It can be argued that the cost of containment for production of pharmaceuticals is relatively small, and while the above systems may have a positive impact from an economic perspective, the cost of confinement would not likely be a key decision point in whether the product was produced. In contrast, however, industrial enzymes have a much lower unit cost, and producers will be much more concerned about the costs of confinement. Perhaps one of the most dramatic examples of this is the enzymes used in the conversion of cellulose from corn stover or other lignocellulosic feedstocks into ethanol. In this case, vast quantities of enzyme are needed compared to the amount required for a typical pharmaceutical product (Aden et al., 2002), and at a greatly reduced price. This may be one of the best uses of the plant production system as it may allow cost-effective quantities of enzymes at large scales that are not reasonably possible by other production systems. However, the current practices for producing pharmaceutical products in plants under USDA regulatory guidelines will make these products prohibitively expensive. Using the proposed system of production, the example below captures several key features of this concept, including an overall reduction in the cost of the end product, a more efficient utilization of the corn plant that results in a greater conservation of our natural resources, and an easier distinction between food and nonfood uses of plants.
In Case 3 we begin with the Case 2 scenario, which is to have the entire transgenic enzyme crop produced at the interior of a given location associated with producing grain ethanol. But in Case 3, we incorporate several other features that will be beneficial in reducing cost.
The first feature of this proposed system is to have the transgenic cellulase enzyme engineered to express only in the germ fraction rather than the entire plant (Hood and Howard, 2002b). The routine separation of the grain into the endosperm and germ fractions is currently practiced in grain ethanol operations today since ethanol is made predominantly from the endosperm fraction, which contains the starch. Having the transgenic cellulase in the germ fraction allows for a source of enzyme at no extra cost in separation and in a fraction that is not currently used in making grain ethanol. Therefore the appropriate cellulose digesting enzymes can be obtained without affecting revenues from the starch fraction or significantly depleting any protein from the distillers dried grain with solubles fraction that is derived from the unused fraction when making ethanol, and subsequently used as animal feed.
The actual cost of the transgenic enzyme would then be associated with the cost of the normally low value germ fraction that is currently used for animal feed. After extraction of the transgenic enzyme from the germ, it may be possible to use the remaining germ product as animal feed, thereby lowering the cost even more. This assumption will rely on obtaining regulatory approval, but there is no reason to believe that these enzymes would be toxic since they are already used as feed additives for animals, and furthermore they would be denatured during processing. Alternatively, it may be possible to use the ground germ fraction containing the transgenic enzymes directly to digest the stover, thereby saving the cost of extraction.
In addition to the benefits of using the grain more efficiently, other features of this system may prove beneficial. One would be that the stover ethanol facility could be located adjacent to the grain ethanol facility to provide manufacturing synergies for ethanol production. The same corn plant that is used to supply the grain for ethanol production could also be used to supply the stover. The transgenic enzyme collected from the germ fraction would be collected from the most central growing area and used to treat stover from the transgenic fields as well as surrounding locations.
A simplified diagram of whole plant usage is shown in Fig. 2, and the production layout is shown in Fig. 3 using the assumptions listed in Table 2. In this way ethanol from grain and stover and the enzymes to process the stover are all being produced from the same acreage, which allows for synergies in transportation and coordination, as well as for efficient utilization of our natural resources. This is a self-contained system with all the necessary components to produce ethanol and there is no additional input required into this system to provide the raw materials for stover ethanol over what is needed to grow the corn for grain ethanol. The savings in this example therefore go beyond the savings for planting a field that otherwise would remain fallow, including
- a more efficient utilization of raw materials without additional inputs, reducing the environmental impact of growing separate crops for grain ethanol, lignocellulosic ethanol, and enzyme,
- a savings of the highly capital intensive fermentation equipment that would be required for making the vast amounts of enzyme required for the biomass conversion into ethanol;
- increased revenue for growers;
- reduced transportation cost because the grain, stover, and enzymes required for ethanol production are supplied in one central location; and
- lowered unit cost of the enzymes since all unit operations are similar to existing practices for growing commodity crops, except mixing the enzyme fraction with the stover.
The proximity of stover to the ethanol plant will likely be one of the critical driving features because of the relatively low value and high transportation cost of the biomass compared to grain or enzyme costs. Therefore, as long as the acreage needed for enzyme is less than acreage for stover, this model could work. Therefore, expression levels of the transgenic enzyme will be critical. In the example above, a high value for expression was used. However, even at expression levels 10 times lower there would still be enough enzyme made within the dedicated area to supply the stover ethanol facility.
In addition to the corn-produced aprotinin and cellulase examples used above to illustrate typical pharmaceutical and industrial proteins, there are many variations that can be applied to this model, including the following:
- Any crop that is dedicated to an industrial application can be adapted to work with transgenic protein products as long as the crop grown for the industrial application can adequately surround the regulated transgenic crop, as defined by regulatory guidelines.
- Growing several industrial enzymes in the same location can be accommodated when the land dedicated to an industrial process would have a relatively large acreage requirement compared to the acreage needed for transgenic enzymes. Therefore, only a very small portion of the field dedicated to industrial applications would be needed to accommodate several transgenic products. There may be limitations as to the proximity of different industrial crops to each other, but this will be based on the product specifications not on the inadvertent mixing in the food supply.
- In the corn examples, the transgenic enzyme is expressed only in the germ. Having the transgenic enzyme expressed throughout the grain can increase the yield of recombinant protein and decrease the transgenic acres required. This model can easily be adapted to collect transgenic enzyme from the endosperm as well as the germ fraction, resulting in lower acreage requirements for the transgenic field. This would be at the expense of revenue for ethanol, and the trade-off can be evaluated for the best overall economic impact.
- Alternative germplasm that provides greater yields of transgenic enzyme per acre can be used. Higher transgenic protein production has been obtained in alternative germplasm such as high-oil corn lines (Hood et al., 2003). This approach would also lower acreage requirements but may also be at the expense of ethanol production if this comes at the expense of reduced starch.
- Producing the industrial enzyme in the stover as well as the grain can increase the overall amount of enzyme obtained. It may be possible to extract enzyme from stover and create purification synergies with that from the grain-produced enzyme. This would lower the overall acreage requirement without necessarily having an impact on ethanol production, since the grain and stover would be unaffected as they relate to biomass for ethanol.
- Ways to engineer the plant to make it more amenable to the specific industrial process are being explored. This would require that the modified transgenic plants outside the buffer zone that do not contain the enzyme product also be kept separate from commodity grain, assuming it had not been deregulated. Therefore a new buffer zone would be required based on the area needed for the modified crop.
- Some of the new transgenic plants may eventually be deregulated but to date have not undergone a thorough regulatory review. Our model may serve as an early process for production until such crops are deregulated.
|
CONCLUSIONS
|
|---|
Our production model proposes a dedicated growing area for transgenic nonfood products without switching between food and nonfood uses on the same site. In concept, this is similar to what is done for microbial and cell culture production systems. All of the USDA regulatory and containment guidelines can easily be met or exceeded, while at the same time economic benefits result for the producers of these crops. This should provide a financial incentive to the producers and give the public and regulatory agencies a higher degree of confidence in those producing these new types of products. The hope is that public acceptance would improve since the production of this type of nonfood product would be easily distinguished from commodity food crops.
The proposed model for growing regulated transgenic crops does not rule out using existing methods of producing these crops in fields that are rotated with food crops. This scenario can easily coexist with current practices. However, as the acreage of specialized crops increases, the cost and challenges will continue to grow. The examples clearly demonstrate the economic, safety, and environmental advantages if this alternative paradigm were utilized. The examples also illustrate how this system can conserve our natural resources through a more efficient utilization of crop plants that are already grown today without additional inputs.
The proposed production system is more than a theoretical curiosity. This approach to growing regulated transgenic crops is feasible today. The examples illustrate incorporating the current trends in increased grain ethanol production to make this a very cost-efficient proposition. The major factor that is lacking is the coordination of growers and ethanol manufacturers. However, many growers have already entered co-ops that are involved in ethanol production making coordination simpler than would otherwise be expected.
|
NOTES
|
|---|
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.
Received for publication September 20, 2006.
|
REFERENCES
|
|---|
- Aden, A., M. Ruth, K. Ibsen, J. Jechura, K. Neeves, J. Sheehan, and B. Wallace. 2002. Lignocellulosic biomass ethanol process design and economics utilizing co-current dilute acid prehydrolysis and enzymatic hydrolysis for corn stover. Technical report, National Renewable Energy Laboratory (NREL/TP-510-32438). Available at http://www.nrel.gov/docs/fy02osti/32438.pdf (verified 14 Feb. 2007). National Renewable Energy Laboratory, Golden, CO.
- Daniell, H., S.J. Streatfield, and K. Wycoff. 2001. Medical molecular farming: Production of antibodies, edible vaccines, and biopharmaceuticals in plants. Trends Plant Sci. 6:219–226.[CrossRef][ISI][Medline]
- Fischer, R., J. Drossard, S. Schillberg, and N. Emans. 1999. Towards molecular farming in the future: Moving from diagnostic protein and antibody production in microbes to plants. Biotechnol. Appl. Biochem. 30:101–108.[ISI]
- Fischer, R., and N. Emans. 2000. Molecular farming of pharmaceutical proteins. Transgen. Res. 9:279–299.[CrossRef][ISI][Medline]
- Giddings, G. 2001. Transgenic plants as protein factories. Curr. Opin. Biotechnol. 12:450–454.[CrossRef][ISI][Medline]
- Hood, E.E. 2002. From green plants to industrial enzymes. Enzyme Microb. Tech. 30:279–283.[CrossRef][ISI]
- Hood, E.E., and J.A. Howard. 1999. Protein products from transgenic plants. Agro-Food Ind. Hi-Tech 3(10):35–36.
- Hood, E.E., and J.A. Howard (ed.) 2002a. Plants as factories for protein production. Kluwer Academic, Dordrecht, the Netherlands.
- Hood, E., and J. Howard. 2002b. Methods for cost-effective saccharification of lignocellulosic biomass. Patent Serial No. 10/310,292, U.S. Publication No. 2003-0109011-A1.
- Hood, E., J. Howard, and D. Delaney. 2003. Method of increasing heterologous protein expression in plants. Patent Serial No. 10/252,732; 60/324,308. U.S. Publication No. 2003-0145346-A1.
- Howard, J.A., and K.C. Donnelly. 2004. A quantitative safety assessment model for transgenic protein products produced in agricultural crops. J. Agric. Environ. Ethics 17:545–558.[CrossRef][ISI]
- Howard, J.A., and E. Hood. 2005. Bioindustrial and biopharmaceutical products produced in plants. Adv. Agron. 85:91–123.[CrossRef]
- Kusnadi, A.R., Z.L. Nikolov, and J.A. Howard. 1997. Production of recombinant proteins in transgenic plants: Practical considerations. Biotechnol. Bioeng. 56:473–484.[CrossRef][ISI][Medline]
- Ma, J.K.-C., P.M.W. Drake, and P. Christou. 2003. The production of recombinant pharmaceutical proteins in plants. Nat. Rev. Genet. 4:794–805.[CrossRef][ISI][Medline]
- USDA. 2003. Field testing of plants engineered to produce pharmaceutical and industrial compounds. Available at http://www.aphis.usda.gov/brs/pdf/7cfr.pdf (verified 14 Feb. 2007). USDA, Washington, DC.
- USEPA. 1989. Risk assessment guidance for superfund. Vol. 1. Human health evaluation manual (Part A): Interim final. Office of Emergency and Remedial Response, EPA/5409/1-89/002. USEPA, Washington, DC.
- USEPA. 1992. EPA guidelines for exposure assessment. Fed. Regist. 57(104):22888–22938. USEPA, Washington, DC.
TOP
ABSTRACT
INTRODUCTION
