Pharmaceutical Crops Have a Mixed Outlook in California

REVIEW ARTICLE ▼ Pharmaceutical crops have a mixed outlook in California

by Michelle Marvier

Crops are being genetically engineered to produce a wide variety of Wilson USDA-ARS/Doug drugs, vaccines and other pharmaceutical proteins. Although these crops may open the door to less expensive and more-readily available drugs, there is concern regarding the potential for contamination of human food and livestock feed, as well as environmental harm. The outlook for the production of pharmaceutical crops in California currently appears mixed. To date, 18 federal permits for field trials involving pharmaceutical or industrial proteins have been approved in California. However, the state’s farming community and general public have thus far rejected pharmaceutical crop production, and Proponents of crops genetically engineered to express pharmaceutical proteins a handful of local governments have believe that these crops could increase the availability of medicines and vaccines, and lower costs. To date, about two-thirds of pharmaceutical field-trials in the United recently banned the cultivation of States have involved maize, a wind-pollinated species (conventional corn is shown). genetically modified crops, including pharmaceutical crops. In light of animal (often a human) or a bacterium. et al. 2004; Ma et al. 2003; Ma, Barros the many pros and cons, three major The genetically modified crops are then et al. 2005). The potential products of cultivated and harvested. transgenic plants include blood thin- approaches — the precautionary ap- In most cases, the crop-produced ners, hemoglobin, insulin, growth proach, risk analysis and cost-benefit pharmaceutical protein is extracted, hormones, cancer treatments and analysis — could be used to move the purified and possibly modified further contraceptives. Products already in debate about pharmaceutical crops before it is administered to humans or the pipeline include plant-produced livestock. In some instances, however, vaccines for hepatitis-B, cholera, rabies, forward. crops are being engineered so that a HIV, malaria and influenza. One com- vaccine can be delivered through the pany is developing genetically modi- ven science fiction writers did not direct consumption of leaves, fruits or fied maize (corn) to produce lipase, a dream that we would someday use other plant parts, without the cost and digestive enzyme used to treat patients maizeE fields to produce insulin, or rice inconvenience of extracting the pro- with cystic fibrosis. Arthritis and other paddies to grow anticoagulants. Today, teins and delivering them via pills or autoimmune diseases are also targets however, crops are being turned into injections (Sala et al. 2003). for plant-produced vaccines (Sala et factories producing not just food, but al. 2003). Researchers have focused Benefits of pharmaceutical crops drugs, vaccines, enzymes and antibod- on maize, bananas, tomatoes, carrots ies. The first step in using crops to pro- Scientists are drawn to the genetic and lettuce as possible oral-delivery duce pharmaceutically active proteins is engineering of crops as a means of mechanisms for such vaccines because the synthesis or isolation of genes that quickly producing large quantities these foods can be eaten raw, thereby code pharmaceutical proteins, followed of drugs and vaccines, with the hope avoiding the protein denaturing that by the transfer of those genes into the that this technology can reduce costs typically occurs during cooking (Sala DNA of crop plants. These transferred and increase the availability of much- et al. 2003). Producing vaccines in food genes, or “transgenes,” can potentially needed pharmaceuticals (Fischer et plants would eliminate the need for re- come from a different plant species, an al. 2004; Giddings et al. 2000; Horn frigeration, which limits the usefulness

http://CaliforniaAgriculture.ucop.edu • APRIL–JUNE 2007 59 Human error occurs and, frankly, is unavoidable. Jack Kelly Clark/UC Davis Kelly Jack Suzanne Paisley/UC Davis Suzanne Paisley/UC

of certain vaccines in many parts of the world (Walmsley and Arntzen 2000). In some cases, the pharmaceuticals targeted for production in transgenic plants are currently produced by cultures of transgenic animal, bacterial or yeast cells in large vats. Plants are an attractive alternative because they could potentially produce greater yields. This is especially important for monoclonal antibodies (such as etanercept, which is used to treat rheumatoid arthritis) because Whenever pharmaceutical-producing crops are grown outside, it is virtually current production methods can- inevitable that birds, insects and other wildlife will eat them, resulting in unknown not keep up with increasing demand risks to the animals, and that the pollen and seeds will be transported offsite. (Elbehri 2005). Moreover, faster and Left, bees on a corn stalk. Right, a red-winged blackbird. less expensive production could reduce prices for consumers. Another tures of genetically modified cells, these ceutical crops. Due to the energetic major benefit of utilizing plants is cells were kept under strict confine- costs that producing pharmaceutical the reduced risk of disease transmis- ment in laboratories. The production proteins likely entails, it is unlikely sion; there is concern that producing of pharmaceutical proteins in maize or that transgenes coding for pharmaceu- drugs via mammalian cell cultures or rice, on the other hand, will typically tical products would persist for very animal milk could facilitate the move- be done in the field where it will be long within the recipient gene pool. ment of certain viruses to humans. impossible to completely contain the However, even transient transgene flow Despite these potential advantages, crops, transgenes and pharmaceutical could cause problems. For example, if drugs produced by pharmaceutical proteins (Ellstrand 2006). Production transgenic pollen fertilizes seed kernels crops have not yet appeared on the of these crops in contained green- on a nontransgenic maize plant, the U.S. market. Several are currently mak- houses or underground caves has been kernels could produce and contain the ing their way through field and clini- proposed as a potential, albeit far less pharmaceutical protein. Alternatively, if cal trials, and the first drugs derived cost-effective, solution. seeds are left behind in the soil follow- from pharmaceutical crops could be Contamination of food and feed. In ing harvest, “volunteer” pharmaceutical on the market within a few years (Ma, 2002, 130 acres of pharmaceutical maize plants could establish themselves in Chikwamba et al. 2005). were cultivated in the United States in subsequent growing seasons, possibly field trials. Two-thirds of all pharma- in mixture with nonpharmaceutical Containment risks ceutical plantings in the United States crops. Because some pharmaceutical There is a long history of cultivating are maize, but soybeans, rice, potatoes, compounds are effective in very low plants to produce pharmaceutical com- alfalfa, wheat, tobacco and other crops concentrations, even low-level contami- pounds, and at least one-fourth of mod- are also being used. The primary con- nation may pose risks. ern medicines contain plant-derived cern is that the public might someday Transgene escape from food crops. ingredients (Raskin et al. 2002). Some find unwanted medicines in their food Although pharmaceutical crops are still plants that are used for pharmaceutical or in livestock feed (Elbehri 2005; Kirk rarely produced and only under tightly production (such as opium poppies) are et al. 2005; Mascia and Flavell 2004; regulated conditions, there already has also food crops (such as poppy seeds). Peterson and Arntzen 2004). been one revealing case of transgene Despite such precedents from nature, Food can be contaminated when escape involving field trials of pharma- genetically modifying major commod- transgenes are not contained, or if plant ceutical maize in Nebraska and Iowa. ity grains such as maize and rice to products intended only for medici- In November 2002, the U.S. Department produce pharmaceutical proteins has nal uses accidentally comingle with of Agriculture (USDA) discovered that raised new levels of concern and public those headed for our dinner tables. ProdiGene had failed to comply with anxiety (Stewart and McLean 2004). Transgenes can escape when pollen federal regulations in two of its field Although earlier methods of pharma- from pharmaceutical crops drifts into trials, which involved maize genetically ceutical production often involved cul- and fertilizes fields of nonpharma- modified to produce a vaccine that pre-

60 CALIFORNIA AGRICULTURE • VOLUME 61, NUMBER 2 Greenpeace/Laura Lombardi

In 2002, field trials of pharmaceutical maize went awry when the producer failed to destroy volunteer maize during the subsequent growing season. As a result, 500,000 bushels of harvested soybeans were destroyed in Aurora, Nebraska. Greenpeace activists hung a banner on the silo. vents diarrhea in pigs. In both locations, these events is that human error occurs pharmaceutical production. As of ProdiGene failed to destroy volunteer and, frankly, is unavoidable. 2003, over three-quarters of field trials maize plants in the subsequent growing conducted to produce pharmaceuti- Food versus nonfood crops season. In Nebraska, the mistake was cal or industrial proteins (including not discovered until after the volunteer The possible escape of pharmaceuti- fibers, plastics and enzymes) had in- maize had been shredded and mixed cal products from engineered crops volved maize, a wind-pollinated spe- among soybeans that had been subse- into the food supply is of concern to cies (Union of Concerned Scientists quently planted at the site. This meant the promoters of genetic engineering, 2003). Grain crops are favored because that 500,000 bushels of soybeans had as well as environmentalists. For ex- protein yields from the large seeds of to be destroyed. In Iowa, there was no ample, an editorial in the journal Nature maize, rice and barley are typically mixing with soybeans, but 155 acres of Biotechnology offered two suggestions much higher than those obtained maize surrounding the pharmaceutical- that could help industry to avoid fore- from leaves and other vegetative crop test site were destroyed because seeable problems (Editors of Nature parts. In addition, pharmaceutical of possible contamination via pollen Biotechnology 2004). First, engineered proteins can remain stable in dried from volunteer plants. ProdiGene was crops could be geographically isolated grain for several years, compared to fined $300,000 for these violations, and to reduce the chances of contamination the much-reduced stability of these also paid $2.7 million in damages and in the general food supply. For example, same proteins in leaf tissue. Moreover, cleanup costs. pharmaceutical crops might be culti- maize is generally recognized as A half-dozen more examples of hu- vated on islands where the food crop is safe for ingestion by the U.S. Food man error involving other, nonpharma- otherwise absent. Second, the editors and Drug Administration (FDA) and ceutical-producing types of genetically recommended that food crops should therefore can be used as an inactive modified crops were reviewed by not be used to produce pharmaceutical carrier, suitable for drug delivery. Marvier and Van Acker (2005). Since the proteins, and that nonfood crop alterna- Despite these advantages, warnings publication of that paper, Syngenta ad- tives such as tobacco might be a wiser from critics may be having an effect. mitted to accidentally distributing the choice. The National Research Council A growing number of companies are seeds of an unapproved variety of ge- (2004) concurred, stating, “Alternative focusing on tobacco, or even mosses, netically modified insect resistant Bt10 nonfood host organisms should be algae and duckweed, as platforms for maize over a 4-year period (Macilwain sought for genes that code for trans- pharmaceutical production (Fischer et 2005), and traces of Bayer’s Liberty Link genic products that need to be kept out al. 2004). These plants, however, pose 601 herbicide-resistant rice have been of the food supply.” risks of their own that must be con- detected in both U.S. and European Despite the National Research sidered. Algae and duckweed, if cul- long-grain food rice, even though the Council’s recommendations, many tivated, would have greater potential variety was never approved or mar- biotechnology firms are nonetheless than highly domesticated crop species keted (Vogel 2006). The lesson from using food grains as platforms for to escape from cultivation.

http://CaliforniaAgriculture.ucop.edu • APRIL–JUNE 2007 61 Additional routes of exposure Even if the production of phar-

maceutical proteins was limited to ANR Communication Services nonfood crops, potential risks would remain. Pollen, fine particles of leaves that are crushed during harvest, and possibly even runoff water contaminated with proteins from decomposing plants, could expose people and wildlife that live on or near pharmaceutical- producing fields to the transgenic ma- terial. Whenever production occurs outside, birds, insects and other wildlife can consume pharmaceutical crops, regardless of where they are grown or what species they are. Pharmaceutical crops may also affect soils and the com- In California, rice farmers strongly opposed efforts to grow 120 acres of rice genetically munity of soil-dwelling organisms. engineered to produce proteins for two pediatric medicines, fearing that their exports to Such impacts on wildlife and soils Asia would be jeopardized. Above, a California rice farm (not genetically engineered). have received scarce attention from scientists and regulators, but surely intended for use in food from those in- none of these compounds contaminates will vary greatly by variety depend- tended for industry. Rapeseed varieties maize intended for food or feed could be ing on the nature of the protein be- containing high levels of erucic acid are prohibitive. ing produced. One possible strategy segregated from those used to produce In addition to rules governing how to avoid these problems would be to canola oil, which must contain less than pharmaceutical crops are grown, USDA engineer proteins so that they do not 2% erucic acid (Ma, Chikwamba et al. inspectors have publicly announced become biologically active until after 2005). Erucic acid is used to create lubri- that field-test sites of such crops will they are extracted and further pro- cants, coatings and surfactants, but the each be inspected five times during the cessed in a laboratory. regular consumption of large amounts growing season and twice postharvest of erucic acid has been linked to heart (Stewart and Knight 2005). However, Regulatory responses disease in animal studies. Producers of based on an audit that included site Pharmaceutical crop varieties are high-erucic-acid rapeseed must maintain visits to 91 field-test locations in 22 not expected to be deregulated; rather, a minimum 16.4-foot buffer zone around states, the USDA Office of the Inspector it is likely that they will only be pro- their fields and clearly label harvested General found that this level of inspec- duced in field trials as permitted under products. In addition, erucic acid levels tion was not consistently maintained. USDA regulatory oversight. Initially, in canola oil are regularly monitored by The audit report concluded that weak- field-trial applications for pharma- various food inspection agencies. nesses in the regulatory oversight of ceutical crops were treated like those Although this example demonstrates genetically modified crop field-trials for any other regulated, genetically the potential for successful segregation, increase the chance that these crops modified crop. However, the USDA more-stringent protocols would be re- will inadvertently persist in the envi- recently published stricter require- quired to produce pharmaceutical pro- ronment (USDA 2005). Of additional ments specifically designed for plants teins in food crops. In the case of erucic concern, the audit found that, “At the genetically engineered to produce acid, a low level of cross contamination conclusion of the field test, APHIS does pharmaceutical and industrial proteins is acceptable (Bilsborrow et al. 1998), but not require permit holders to report (Federal Register 2003). These new for pharmaceutical compounds there is on the final disposition of genetically requirements aim to reduce the risk generally zero tolerance. Studies exam- modified pharmaceutical and industrial of gene flow and the contamination of ining the potential for the coexistence of harvests. . . . As a result, [the inspectors] food and feed. Confinement measures other types of genetically modified crops found two large harvests of genetically now required for pharmaceutical crops with nongenetically modified varieties modified pharmaceutical crops remain- include greater geographic isolation demonstrate that contamination can be ing in storage at the field-test sites for over from other fields of the same species, limited (for example, less than 0.9%) but a year without APHIS’s knowledge or buffer zones of bare soil around the not entirely prevented (EuropaBio 2006). approval of the storage facility” (USDA field edge, scouting for and destroying Moreover, in the rapeseed example, only 2005). Clearly, better adherence to moni- volunteer plants in subsequent field one or two compounds must be monitoring requirements is needed to mini- seasons, and the dedication of equip- tored. In contrast, if maize is eventually mize the risk of a loss of containment. ment for use only on the trial fields. used to produce some 50 or 100 different Although the 2003 regulations set There is a precedent for the suc- pharmaceutical compounds, the costs forth by USDA are an important step, cessful segregation of crop varieties for systematic monitoring to ensure that the proposed rules make no attempt

62 CALIFORNIA AGRICULTURE • VOLUME 61, NUMBER 2 to protect wildlife (fencing or netting exactly how many field trials of phar- Field-testing in California are not required), assess how pollen maceutical crops have been approved or fine particulate matter from the The USDA database of field-trial in California. However, the Union of crop might affect humans, or test soils permits for plants expressing phar- Concerned Scientists (2007) estimates and groundwater for pharmaceutical maceutical and industrial proteins that 18 permits for field trials involving residues. Also missing is any require- includes many entries for which the pharmaceutical or industrial proteins ment that the pharmaceutical variety petitioning organization has used were approved in California, the earli- be readily identifiable. For example, a claim of Confidential Business est in 1996 and one as recently as 2006 several authors have suggested that Information to withhold from the (table 1). According to this analysis, pharmaceuticals could be produced public any information regarding California is tied with Kentucky for in “identity-preserved varieties, such the transgene, its source or the traits seventh among U.S. states and territo- as white tomatoes or maize, which are that have been altered (USDA APHIS ries, after Nebraska with 41 approved easily identified by their pigmentation” 2007). It is therefore difficult to know permits, Hawaii with 40, Puerto Rico (Ma et al. 2003).

No specific requirements were TABLE 1. USDA-approved field-trial permits allowing the growth of crops genetically engineered proposed for molecular solutions to to produce pharmaceutical or industrial proteins in California, 1996–2006 contamination, presumably because these are not sufficiently developed Engineered Issued/ Source of yet. However, molecular strategies hold crop Applicant effective gene* Pharmaceutical or industrial protein great promise for the improved contain- Dow 6/2002 CBI† CBI: Unidentified pharmaceutical protein ment of transgenes. Examples include CBI: Unidentified transcriptional activator 3/2001 CBI the genetic modification of chloroplast (pharmaceutical) Monsanto DNA rather than nuclear DNA (for crop CBI: Unidentified transcriptional activator 3/2001 CBI species in which pollen does not con- (pharmaceutical) tain chloroplasts, transgenes would not CBI: Unidentified novel protein that may 3/2000 Unclear‡ have pharmaceutical or industrial uses move with pollen) (Daniell et al. 2002) Maize and the inducible production of phar- CBI: Unidentified novel protein that may 4/2001 Unclear maceuticals (for example, the pharma- have pharmaceutical or industrial uses ceutical protein is activated by exposure Pioneer CBI: Unidentified industrial enzyme and to ethanol vapor) (Mascia and Flavell 4/2002 Unclear unidentified novel protein that may have pharmaceutical or industrial uses 2004). The tissue-specific expression CBI: Unidentified novel protein that may of pharmaceutical proteins may also 4/2004 Unclear have pharmaceutical or industrial uses reduce or eliminate certain avenues of Leaf USDA 3/2004 Unclear CBI: Unidentified industrial enzyme exposure (such as the possibility of ex- mustard Agricultural posure via pollen inhalation), and gene Research deletion technologies could potentially CBI Service 3/2004 CBI CBI: Unidentified industrial enzyme be used to remove transgenes from Rapeseed Pioneer 9/1996 CBI CBI: Unidentified pharmaceutical protein certain tissues (such as pollen) to re- Pharmaceutical proteins: Antithrombin and 3/1997 Humans duce the possibility of transgene spread serum albumin (Keenan and Stemmer 2002). Pharmaceutical proteins: Antitrypsin, 2/1998 Humans If transgenes could be contained, antithrombin and serum albumin then regulations could be much more 2/1998 CBI CBI: Unidentified pharmaceutical protein Ventria permissive about which traits are al- CBI: Unidentified pharmaceutical protein Bioscience 5/2000 CBI and unidentified novel protein that may lowed in crop plants. On the other Rice (formerly have pharmaceutical or industrial uses hand, if transgenes will inevitably Applied Phytologics) Pharmaceutical proteins: Antitrypsin, escape and spread — despite our best 4/2001 Humans lactoferrin and lysozyme intentions for containment — then we Pharmaceutical proteins: Lactoferrin must be much more cautious about 4/2003 Humans and lysozyme which traits are allowed to be devel- Pharmaceutical proteins: Lactoferrin 5/2004 Humans oped in crop plants. Alternatively, the and lysozyme cultivation of crops engineered to pro- Mice, Planet Antibodies to tooth decay and Tobacco 6/2006 rabbits, duce particularly hazardous pharma- Biotechnology common cold ceutical proteins might be restricted CBI to greenhouses or other enclosed * Refers specifically to the gene coding for the industrial or pharmaceutical protein. † CBI = Confidential Business Information. facilities, such as caves. Although pro- ‡ Source of gene coding for industrial and/or pharmaceutical protein(s) cannot be determined from publicly available duction in such facilities is feasible, information. it would likely be far more expensive Source: Union of Concerned Scientists 2007. than field production.

http://CaliforniaAgriculture.ucop.edu • APRIL–JUNE 2007 63 with 39, Wisconsin with 29, Iowa with 27 and Illinois with 19. Pharmaceutical rice. The production of pharmaceutical proteins in transgenic crops is meeting with some resistance in California, as Ventria Bioscience recently discovered. Ventria had received federal Campbell Communications/Tom Agricultural Purdue permits to grow approximately 100 acres of pharmaceutical rice in California almost annually since 1997 (see table 1). However, the company’s plans to expand its 2004 field trials to 120 acres of rice engineered with synthetic human genes were met with strong opposition from California rice farmers and environmentalists. Ventria’s rice has been genetically engineered to produce lactoferrin and lysozyme, compounds used to treat severe diarrhea in infants. However, farmers were concerned that even low levels of In an abandoned Indiana mine, Controlled Pharming Ventures is working contamination of their rice crops could with Purdue University researchers to develop techniques for growing threaten exports to Asia. pharmaceutical crops underground, in order to limit risks. The California Rice Certification Act of 2000 gave the California Rice crops outright, and in some cases bans concerns for the environment or even Commission the authority to devise have indeed been implemented. Bans food safety. The contamination of U.S.- protocols governing the cultivation on all genetically modified plants are produced rice with the unapproved of any new rice variety that requires in effect in four counties: Mendocino Liberty Link 601 (herbicide-resistant) segregation. Despite farmers’ concerns, (Measure H, passed by voters in March variety has had an enormous economic on March 29, 2004, the commission ap- 2004), Trinity (passed by the county impact on U.S. rice growers. U.S. export- proved planting guidelines for Ventria’s board of supervisors in August 2004), ers of long-grain rice lost about expanded plantings in a 6 to 5 vote, on Marin (Measure B, passed by voters $150 million because genetically modi- the condition that the field trials be con- in November 2004) and Santa Cruz fied rice is banned throughout most of ducted in counties such as Orange and (unanimously passed by the county the European Union, a major importer San Diego, remote from the state’s rice- board of supervisors in June 2006). of U.S. long-grain rice. Even greater growing regions. Due to the late timing In contrast, voters rejected initiatives economic losses would likely occur if of the commission’s decision and the to ban genetically modified crops in a crop were found to be contaminated need to plant immediately, Ventria then four counties: Humboldt, San Luis with a pharmaceutical protein. Whether asked the California Department of Obispo and Butte in 2004, and Sonoma pharmaceutical-producing crops will Food and Agriculture (CDFA) to issue in 2005. Supervisors in several other be accepted in California will likely an emergency permit for the proposed California counties, including Fresno, depend on the economic value of other field trials. On April 9, 2004, CDFA de- Kern and Kings, have passed resolu- markets that might be placed at risk. cided not to approve Ventria’s proposal tions supporting the use of genetically A proposal to produce pharmaceutical because federal regulators at USDA modified crops. rice within a major rice-producing area had not yet completed their review of The political future of local measures, such as the Sacramento Valley is un- Ventria’s permit application. California either for or against genetically engi- likely to be welcomed. However, a pro- regulators essentially deferred to fed- neered crops, was recently challenged by posal to grow that same pharmaceutical eral regulation, reasoning that federal Senate Bill 1056, which would have pro- rice in an area with very little other rice oversight of the field-trial application is hibited California counties, towns and production may be acceptable. both necessary and sufficient. In 2005, cities from passing any local regulation Evaluating risks and benefits Ventria attempted to move its field trials of seeds and nursery plants. However, in to Missouri, where it met similar resis- September 2006, this bill failed to make All forms of agriculture entail some tance from major rice purchasers. it out of committee and died with the risks to the environment. Whenever Local bans. Although California reg- close of the legislative year. The failure of food is grown, some species lose their ulators may be happy to defer to USDA this bill leaves open the possibility of ad- habitat and some may be poisoned, judgment when it comes to genetically ditional local restrictions on genetically trapped or shot; species extinctions are modified crops, the public and local modified crops in the future. also possible. Pharmaceutical crops communities are not always so accom- Economic considerations. In the end, entail all of these same risks plus addi- modating. Several counties have con- economic concerns regarding the con- tional ones — the contamination of food sidered banning genetically modified tainment of food crops may outweigh and feed being the most serious. There

64 CALIFORNIA AGRICULTURE • VOLUME 61, NUMBER 2 (2) Formal risk-assessment framework. The U.S. Environmental Protection Agency embraces a risk- assessment approach in all of its USDA/Ken Hammond USDA/Ken regulatory capacities, including the regulation of chemical pesticides and “biopesticides,” such as plants genetically modified to express insecticidal proteins. Risk is defined as a function of both hazard and exposure, such that either a low hazard or low probability of exposure will reduce the assessed level of risk. Hazard is a measure of the harmful effects of the pharmaceutical proteins on people and the environment; as such, not all are equally hazardous. For example, lactoferrin is naturally produced in human tears and breast milk. Assuming that plant- produced lactoferrin is very similar to human-produced lactoferrin, this compound would present little if any hazard to humans. With regard to exposure, the potential routes and amounts of exposure to pharmaceutical compounds are expected to be highly variable. Exposure will depend upon which crop species is used as the production platform, where it is grown, and where the protein is and is not expressed within the plant (pollen, for example, is highly mobile). The amount of land needed to produce sufficient quantities of In a 2004 report, an expert panel of the National Research Council recommended that particular pharmaceuticals must also food crops should not be used to produce pharmaceutical crops, suggesting instead that nonfood crops such as tobacco (shown in Virginia) would be a wiser choice. be considered; this will depend both upon demand for the product and the protein yields obtained per plant. are three major approaches to evaluating However, interpretations of the pre- Incorporating transgenes into chloro- the potential benefits and risks. cautionary approach vary. A strong in- plast DNA rather than nuclear DNA (1) Precautionary approach. A pre- terpretation mandates that the producer could reduce exposure both by limit- cautionary approach typically shifts demonstrate the absence of harmful ing the expression of the protein in the burden of proof onto the producer, effects prior to the release of the prod- pollen and by boosting the production so that a practice or product is not uct. Given that harmful effects could of target proteins to a level where suf- approved until there is sufficient sci- be exceedingly rare, this represents an ficient quantities could be produced in entific understanding of the potential impossible standard from a scientific very small fields (Daniell et al. 2002). risks. This approach has been adopted perspective. In contrast, a weak in- (3) Cost-benefit analysis. An im- in many legal and policy arenas, interpretation mandates that regulators portant component of the cost-benefit cluding the transnational movement should only consider delaying the ap- analysis approach is “fairness” — who of living, genetically modified organ- proval of a practice or product when benefits and who pays the costs. Fairness isms under the Cartagena Protocol on sufficient evidence of risk exists (Conko is a core value of many Americans, Biosafety. Since all nations with com- 2003). California counties with morato- and environmental policy discussions mercial transgenic production must ria on all transgenic crops are adopting increasingly focus on equitability and undertake safety testing (of some sort) a strong interpretation of the precau- fairness. The precautionary approach prior to the commercial production of tionary approach, similar to European and risk-assessment framework do not transgenic crops, a precautionary countries that require the labeling of require the consideration of costs and approach is already being applied to a any foods with genetically modified benefits to stakeholder groups. But one certain degree (Conko 2003). plant ingredients. explanation for the public’s reluctance

http://CaliforniaAgriculture.ucop.edu • APRIL–JUNE 2007 65 regarding the production of pharmaceu- the research and development of ise. Bananas that could cheaply and easily tical proteins in crop plants could be that pharmaceutical crops will likely re- deliver vaccines to children throughout the distribution of benefits (primarily to main very expensive. the tropics could be a wonderful inven- corporations) does not match the distri- Other potential benefits are possi- tion. But there are downsides; it will be bution of risks (primarily falling upon bly increased income for farmers and difficult to avoid food contamination and the general public). higher tax revenues (Wisner 2005). potential harmful effects to wildlife if Because biotech and pharmaceuti- There is much hope that pharmaceu- pharmaceuticals are widely produced in cal companies are the primary eco- tical crops will improve farmer in- food crops grown out of doors. nomic beneficiaries, the key questions comes, but these benefits are unlikely Finally, the pros and cons of alterna- for a cost-benefit approach applied in a global market where the produc- tive strategies to achieve the same goals to pharmaceutical crop production tion of pharmaceutical proteins in should be assessed (O’Brien 2000). For are whether the economic rewards genetically modified crops could be example, could certain pharmaceuti- outweigh the potential risks of un- undertaken in whichever nation has cal crops reasonably be confined to wanted pharmaceutical exposure, the lowest production costs and weak- greenhouses, caves or other enclosed and whether the distribution of the est regulatory restrictions (Wisner facilities? Are there other possible costs and benefits is equitable and 2005). Another important issue for routes to the inexpensive and efficient fair (Elbehri 2005). If economic profits farmers concerns liability for contami- production of drugs that perhaps do are reinvested into the research and nation incidents. In the only precedent not involve the transgenic manipula- development of new drugs, then addi- to date, ProdiGene was held account- tion of crop plants? The future course of tional benefits for human and animal able for its mistakes. Communities this technology will require thoughtful health may be achieved. In addition, or regulatory agencies considering input from ecologists, public health ex- drug prices might be reduced if it allowing the production of pharma- perts and medical researchers — as well becomes inexpensive to manufacture ceutical crops will want assurances as those who genetically engineer these drugs in large quantities. However, regarding who pays for any damages. crops in the first place. because most pharmaceutical crops A promising new technology? are designed to produce patented pharmaceutical compounds, there Like many new technologies, the M. Marvier is Associate Professor, Environmental would typically be little competition genetic engineering of crops to produce Studies Institute and Department of Biology, to drive prices lower. Furthermore, pharmaceutical products has great prom- Santa Clara University.

References Bilsborrow PE, Evans EJ, Bowman J, et al. 1998. Kirk DD, McIntosh K, Walmsley AM, et al. 2005. next generation of U.S. agricultural biotechnology: Contamination of edible double-low oilseed rape crops Risk analysis for plant-made vaccines. Transgen Res Politics, policy, and plant-made pharmaceuticals. Tech- via pollen transfer from high erucic cultivars. J Sci Food 14:449–62. nol Forecast Soc 72:521–34. Agric 76:17–22. Ma JK-C, Barros E, Bock R, et al. 2005. Molecu- Stewart PA, McLean W. 2004. Fear and hope Conko G. 2003. Safety, risk and the precaution- lar farming for new drugs and vaccines. EMBO Rep over the third generation of agricultural biotechnol- ary principle: Rethinking precautionary approaches 6:593–9. ogy: Analysis of public response in the Federal to the regulation of transgenic plants. Transgen Res Register. AgBioForum 7:133–41. Ma JK-C, Chikwamba R, Sparrow P, et al. 2005. 12:639–47. Plant-derived pharmaceuticals — the road forward. Union of Concerned Scientists. 2003. Pharm Daniell H, Khan MS, Allison L. 2002. Milestones in Trend Plant Sci 10:580–5. and Industrial Crops: The Next Wave of Agricultural chloroplast genetic engineering: An environmentally Biotechnology. Cambridge, MA. www.ucsusa.org/ Ma JK-C, Drake PMW, Christou P. 2003. The produc- friendly era in biotechnology. Trend Plant Sci 7:84–91. food_and_environment/genetic_engineering/ tion of recombinant pharmaceutical proteins in plants. pharm-and-industrial-crops.html. Editors of Nature Biotechnology. 2004. Drugs in Nat Rev Genet 4:794–805. crops — the unpalatable truth. Nat Biotechnol 22:133. Union of Concerned Scientists. 2007. Pharma Macilwain C. 2005. US launches probe into sales of Crop Approvals in the United States. Cambridge, Elbehri A. 2005. Biopharming and the food sys- unapproved transgenic corn. Nature 434:423. tem: Examining the potential benefits and risks. MA. http://go.ucsusa.org/food_and_environment/ AgBioForum 8:18–25. Marvier M, Van Acker R. 2005. Can crop transgenes pharm/ (accessed Feb.18, 2007). be kept on a leash? Front Ecol Environ 3:99–106. Ellstrand NC. 2006. When crop transgenes wander [USDA] US Department of Agriculture. 2005. in California, should we worry? Cal Ag 60(3):116–25. Mascia PN, Flavell RB. 2004. Safe and acceptable Audit Report: Animal and Plant Health Inspection strategies for producing foreign molecules in plants. Service Controls Over Issuance of Genetically En- EuropaBio. 2006. Understanding coexistence: Sci- Curr Opin Plant Biol 7:189–95. gineered Organism Release Permits. Office of the ence, principles and practical experience. ABE/Europa- Inspector General. Audit #50601-8-Te, December. Bio. www.europa-bio.be/documents/040406/ National Research Council. 2004. Biological Confine- Understanding%20Coexistence%20Fact%20File.pdf. ment of Genetically Engineered Organisms. Washington, [USDA APHIS] Animal and Plant Health Inspection DC: Nat Acad Pr. 284 p. Service. 2007. USDA Release Permits for Pharmaceuti- Federal Register. 2003. Field testing of plants cals, Industrials, Value Added Proteins for Human Con- O’Brien M. 2000. Making Better Environmental Deci- engineered to produce pharmaceutical and industrial sumption, or for Phytoremediation Granted or Pending sions: An Alternative to Risk Assessment. Cambridge, compounds. Fed Reg 68:11337–40. by APHIS. www.aphis.usda.gov/brs/ph_permits.html. MA: MIT Pr. 352 p. Fischer R, Stoger E, Schillberg S, et al. 2004. Plant- Vogel G. 2006. Tracing the transatlantic spread Peterson RKD, Arntzen CJ. 2004. On risk and plant- based production of biopharmaceuticals. Curr Opin of GM rice. Science 313:1714. Plant Biol 7:152–8. based biopharmaceuticals. Trend Biotechnol 22:64–6. Walmsley AM, Arntzen CJ. 2000. Plants for Giddings G, Allison G, Brooks D, et al. 2000. Trans- Raskin I, Ribnicky DM, Komarnytsky S, et al. 2002. delivery of edible vaccines. Curr Opin Biotechnol genic plants as factories for biopharmaceuticals. Nat Plants and human health in the twenty-first century. 11:126–9. Biotechnol 18:1151–5. Trend Biotechnol 20:522–31. Wisner R. 2005. The Economics of Pharmaceuti- Horn ME, Woodward SL, Howard JA. 2004. Plant Sala F, Rigano MM, Barbante A, et al. 2003. cal Crops: Potential Benefits and Risks for Farmers molecular farming: Systems and products. Plant Cell Vaccine antigen production in transgenic plants: and Rural Communities. Union of Concerned Scien- Rep 22:711–20. Strategies, gene constructs and perspectives. Vaccine tists, Cambridge, MA. www.ucsusa.org/food_and_ Keenan RJ, Stemmer WPC. 2002. Nontransgenic 21:803–8. environment/genetic_engineering/ crops from transgenic plants. Nat Biotechnol 20:215–6. Stewart PA, Knight AJ. 2005. Trends affecting the economics-of-pharmaceutical-crops.html.

66 CALIFORNIA AGRICULTURE • VOLUME 61, NUMBER 2