October 2000 Volume 18 Number Supp pp IT59 - IT61


Agricultural biotechnology

Technological progress has been rapid, but companies can no longer afford to ignore public concerns.



It has been said that the genetic manipulation of plants is the reason why the genomics revolution—touted as the third technological revolution following the industrial revolution and the computer revolution—will have a major global impact1. Genomics will certainly influence new drug discovery to treat human diseases, but through its application to agricultural biotechnology, a proportion of our needs for fuel, fiber, food, and some medicines will one day be obtained from genetically modified plants.


Global sales from transgenic crops were estimated at $1.6 billion in 1998, and rose by more than a third to an estimated $2.1 to $2.3 billion in 1999, according to the International Service for the Acquisition of Agri-Biotech Applications (ISAAA)2. It is not surprising, therefore, that with such a large market, agricultural biotechnology is a centerpiece of major industrial efforts. Large chemical and pharmaceutical companies, including Rohm and Haas, Dow Chemical, DuPont, Monsanto, and Novartis have invested heavily in agbiotech—specifically, genetic engineering of plants—both in-house and through major collaborations with genomics companies.


Historical perspective

Numerous developments over a century or more have contributed to the current progress of agricultural biotechnology and transgenic plants. One focus of current plant genetic engineering is the use of plant viruses, such as tobacco mosaic virus (TMV), to carry foreign genes and ensure their expression in the plant. The existence of plant viruses was postulated in 1898 by Beijerinck, but did not gain wide acceptance until much later. First isolated in the form of particles of an enzyme-like protein in 1935, TMV was characterized as a nucleoprotein in 1937 (reviewed in ref. 3). Research on TMV's structural aspects continued through the 1960s4, and by the mid-1980s it was shown that tobacco plants could be genetically transformed to resist virus disease development, for example, by expressing the TMV coat protein5.

                                                                                                                                                  This work was aided by the development of another key tool in plant transgenics: the Ti plasmid of Agrobacterium tumefaciens, used as a workhorse of plant genetic engineering to shuttle foreign genes into plant cells. Several other approaches for delivering DNA to plant cells were also developed, including chemical methods and electroporation, microinjection, and ballistic methods. As monocotyledonous plants are generally not amenable to transformation by Agrobacterium, these methods were particularly important for facilitating stable gene transfer to many of the major monocot crops.


The use of tobacco plants and TMV as a platform on which to try different approaches to the genetic engineering of desirable traits has continued unabated. By the late 1980s,  plants were being engineered that expressed antisense RNA against TMV coat protein and were resistant to disease progression6. Also, the first monoclonal antibodies (Mabs) were produced in transgenic plants, paving the way for a low-cost, high-yield alternative for Mab production7, and confirming the potential of plants as carriers for the production of novel materials. This and other work has led to experiments to produce textile fibers, fuel oil, plastics, vaccines, nonplant enzymes, and other materials in plants. Transgenic varieties of major crops, such as corn, soybean, cotton, and canola were first planted on a commercial scale in the past decade, and Calgene's (Davis, CA) groundbreaking Flavr-Savr transgenic tomato was introduced in 1996, albeit only for a short period of time.


Current state

According to ISAAA, the world market for genetically engineered plants will be $3 billion in 2000, $8 billion in 2005, and $25 billion by 2010. Transgenic plants will take the majority of the largest crop seed markets, and will be planted on about 60 million hectares annually, mostly in the US2. Herbicide-resistant soybeans and insect-resistant corn and cotton have already proved to be agriculturally successful products, both in the field and on the bottom line of the agrochemical companies that make the seeds.

The large agrochemical and pharmaceutical companies that seek to exploit the vast and lucrative agbiotech markets have realized that the key technological driver behind transgenic plants is genomics. Genomics arose as part of the effort to sequence the entire human genome. The efficient DNA-sequencing capabilities and gene expression assays it spawned are traditionally used to screen for drugs for human use, but these companies are now focusing these technologies on the identification of novel plant and plant pathogen targets, against which they can screen their vast compound libraries. In order to achieve these aims, they are turning their attention to biotech companies oriented toward genomics, combinatorics, and bioinformatics, whose primary interest is not agbiotech, but whose tools can find a ready application there.


Table 1 lists some recent deals involving large agrochemical or pharmaceutical companies and biotech companies. It can be seen that agbiotech deals can be very substantial, as illustrated by the acquisitions of DeKalb Genetics and Plant Genetic Systems and the Monsanto/Millennium collaboration, and that the whole gamut of high-throughput target and lead identification and validation are involved. This is very important from the perspective of smaller biotech companies whose technologies may be suffering from too much competition in the human healthcare arena—their technologies can be easily applied to the agbiotech field. Table 2 lists the genomic and screening efforts of some of the largest agrochemical and pharmaceutical companies.


On the technology front, present experience collected from transgenic plants suggests that simplicity in attempting to engineer a desired trait is the key factor for success. For example, introducing just one or a few foreign genes into a plant, with minimal effects on its physiology, is the best way forward. A case in point is the engineering of glyphosate herbicide resistance in plants. Glyphosate is a major herbicide, and achieving resistance to it in the desired plant crop enhances the herbicide's selectivity. Glyphosate-resistant plants are created by engineering intothem a single bacterial enzyme that is highly resistant to the herbicide.


Another example is the engineering of insect-resistant plants by the addition of the gene for one of the insecticidal toxins of Bacillus thuringiensis (Bt). The financial returns that result from these simple protocols are significant. For example, in the US in 1997, transgenic corn expressing a Bt toxin had a 7% increase in yield per acre and an increased net return per acre of $16.888.


Industry challenges

While advances in genetic engineering have enhanced our ability to manipulate the plant genometo achieve desirable traits, challenges remain. For example, it cannot be predicted a priori whether the best way to engineer a trait into a plant is conventional breeding and transgenic methodology or the use of plant virus gene vectors. Virus-based transient RNA and DNA gene expression is rapid, convenient, and widely applicable, whereas conventional crossbreeding is not. However, instability of the foreign gene in the viral genome can present problems, and expression efficiency may not always be under tight control. However, a great deal of work is being done to improve plant viruses as the main tools for the engineering of transgenic plants, including gene replacement, gene insertion, antisense approaches, epitope presentation, use of gene expression cassettes, complementation, and others9.


Another technical challenge is that, irrespective of the mode of novel gene introduction, it is not sufficient merely to introduce a gene into a plant. The gene should exist in single copy, and the vector carrier should not be integrated into the plant genome. In addition, it is necessary to be able to predict the quantitative expression of transgenes to ensure that the traits do not "jump" between species, and also that no unwanted gene expression occurs10. These challenges are akin to the ones faced by human gene therapy, which is confronted with the requirement to deliver and express genes very precisely and in a controlled manner. In both instances, these are issues of which the industry is well aware, and concerted efforts are being made to address them.


The future

Transgenic plants are being engineered with a variety of useful traits that do not always fall within the typical categories of higher yield, insect or herbicide resistance, longer shelf-life, and the like. A particularly interesting application of transgenic plants to look out for is their use for bioremediation, specifically the reclaiming of metal-contaminated soils.

Phytoremediation is a potentially cheap way to achieve soil remediation, and it may even be possible to recycle metals from these plants. For example, plants are often able to accumulate metals such as selenium and mercury and tolerate them, both of which are important traits if they are to be used as bioremediators. In addition, soil lead and chromium may be inactivated in the soil itself by plants. However, there is at present little understanding of the molecular mechanisms by which plants achieve these effects naturally. Progress has been reported recently on the uptake of iron, zinc, and cadmium by Arabidopsis and yeast mutants, and this information will be used in the future to develop commercially viable plants used for metal phytoremediation11.


The future will see continued efforts to use transgenic plants as factories for vaccine production, including even immunocontraceptive vaccines, something that is extremely important in the context of eradicating disease from the developing world while also providing food12. In addition, it is now possible to engineer the expression of antibodies in plants to specific requirements. Not only do plants offer unique alternatives for bulk production of antibody molecules, but they are also being engineered to assemble full-length and complex, multimeric antibodies, overcoming some of the original limitations associated with assembly and glycosylation patterns13.


Another interesting prospect is the increasing availability of plant genomic knowledge. As plant genomes begin to become fully known, the same types of advances that knowledge of the genomes of human or other organisms have led to will be enabled in plants as well. For example,  the variation in plant microsatellite DNA is now being explored in a systematic fashion14. Understanding this variability and its implications may lead to the plant equivalent of pharmacogenomics, whereby genetic variation in a plant population is exploited in order to target specifically certain treatments, or where this variation leads to a better understanding of the underlying molecular mechanism of plant disease.


The future will also see the increasing convergence of cell signaling research, which is a major area with very significant impact in human drug discovery efforts, to plants. For example, a recent report describes how a significant control component of the growth of meristem stem cells is a signaling pathway that involves a CLV1/CLV2 receptor kinase complex, and that this CLV pathway represses the activity of the transcription factor WUSCHEL and thus effects its control15. At the same time, there is constant development of novel assays for plant receptor kinase activation and signaling mechanisms16. This work will eventually lead to the discovery of novel ligands against key plant signaling proteins, which itself will enable the development of much more specific pesticides and herbicides.


In addition to cell signaling research, the future will also see the application of other mammalian molecular biology methods to plants, including antisense. A recent study used antisense approaches to show the involvement of GTPases in plant defense reactions17, and this work will also eventually lead to a better understanding of plant physiology at the molecular level.


Finally, the future will also see the development of several novel insecticidal toxins expressed in new transgenics that do not originate from Bacillus thuringiensis. One recent report describes such toxins cloned from Photorhabdus luminescens and from Xenorhabdus nematophilus18. These are significant developments of alternative highly selective insecticides for the future.



It may be surprising to know that at present, more than half of the world's soybean crop and about one-third of the corn crop is transgenic. Soybeans and corn are part of hundreds of everyday foods such as cereals, cooking oils, corn syrup, soft drinks, and sweets. Nevertheless, the general public has considerable reservations about transgenic plants. In Europe, attitudes are much more negative than in the US, and a string of US–European Union reciprocal disputes and trade blocks highlight the urgent need to ensure the need for widespread education about the technology, its applications, and its implications.


If agbiotech is to realize its full potential, its developers and practitioners will need to make an even more concerted effort to share their knowledge with the public, showing how all relevant concerns are being addressed. Some major public concerns include the segregation of genetically modified (GM) and non-GM crops, the potential cross-pollination between wild species and GM crops, uses of marker genes, the potential of having new allergens introduced into the food chain, and the actual safety of GM foods. These valid and important considerations highlight how achieving a balance of technological progress, extensive and appropriate oversight, and public consultation will ensure the ultimate success of what is a major technology revolution.


Reprinted from Nature Biotechnology 17, 612–614 (1999).