Use of Experimentation Techniques in Biotechnology

Use of Experimentation Techniques in Biotechnology!

Field testing of genetically engineered plants has come a long way since the first attempts of 1985-1986. It has come to be seen as a fully integrated part of all the projects that aim at eventual commercial introduction of such plants, and as an important component of many fundamental research projects in plant biology.

Field experiments with genetically engineered plants are not a goal in themselves but a means to obtain information which should allow the performer of the experiment to make some progress in his or her project. It is therefore the nature of the project in the context of which a field release is carried out that will determine the way in which a trial is organised.

The variety of projects that include field trials with transgenic plants has grown substantially, and covers very diverse areas of fundamental and applied research in biology and agronomy. The ultimate goals of such projects are also very different.

Some projects aim at the production of new crop varieties incorporating new genes introduced by genetic engineering. These are the most widely publicised insect, virus or herbicide resistant plants, plants with altered protein or lipid composition, etc.

Most of these projects are executed by industrial corporations, sometimes in collaboration with public research institutes. They also include many of the earliest achievements in plant genetic engineering.

Another rapidly growing group of projects involves the creation of transgenic plants, and the testing of them in some environments is only a step on the road to better understanding the processes governing the development of plants and their interaction with the outside world. In short, genetically engineered plants are generated with one of two broad goals in mind: the production of material products, or the production of knowledge.

Field Trials in Plant Biotechnology:

Testing in the field is usually part of a more extensive sequence of tests performed on plants. The relative importance of field tests in such a sequence is determined by the nature of the trait which has been introduced, by the plant species under investigation, and by the ultimate goal of the project.

If the goal of the project is to test the phenotypic expression of a gene involved in some aspect of the plant metabolism, most of the work may actually be done in the laboratory, in growth chambers, or in a greenhouse. This often has the advantage of allowing better control over the environmental conditions in which the plant is tested, resulting in more clear-cut experimental results.

In such projects, scientists may actually avoid bringing plants outdoors, since it does not help them in their work. Hence, an important part of in vivo experimentation in plant physiology, phytopathology and even ecology takes place in a controlled environment. However, a controlled environment only allows the testing of those interactions between plants and their environment which are allowed to vary.

That is why no serious model experiment can be interpreted satisfactorily without reference to those conditions which are kept constant, and without considering possible other interactions in an open field situation, regardless of the nature of the plants being studied.

It explains why, after nearly a century of research in controlled environments, nearly all the research in plant breeding and agronomy is still done in the field. There is simply no other effective way to integrate all the variable components of the environment when testing a trait expressed in a plant line for its genetic or agronomic potential.

The same applies for genetically engineered plants, even for “simple” characteristics such as monogenic resistance to pests, diseases or herbicides. The detailed reasons for the necessity of testing transgenic lines in the field will be further discussed later, with a number of practical examples. Now, the practical implications of the existing regulatory framework for field testing genetically engineered plants, as compared with other plants.

Regulatory Constraints on Field Tests:

It is important to be aware of the fact that agronomic or genetic field testing is a matter of statistics, of working with quantitatively changing parameters. A plant is not either disease resistant or susceptible: it is scored on a continuous scale from completely devastated to (sometimes) almost completely unaffected. The amount of damage varies depending on a wide range of factors which at first sight may seem to have little to do with either the plant or the pest attacking it.

Therefore any plant breeding programmes which includes selection for characters such as disease resistance requires very extensive testing programmes under a wide range of natural environments, usually over several years.

The same is true of resistances built into the plant by genetic engineering. However, there are major differences in the way experiments with genetically engineered plants are set up, compared with non-engineered plants.

1. Field tests with genetically engineered plants require a permit. Trivial as this may seem, it is a very major new situation for scientists working in agronomy and plant breeding, and it takes time for the research community to adjust. The process of obtaining such a permit is complicated and time-consuming. Moreover, the data required in the application for permits is often quite difficult to find, making the preparation of the application a significant part of the experimental work as a whole.
2. Regulations usually require that experiments with transgenic plants be set up in conditions of isolation from related plants, in order to minimise the chance that the engineered genes “escape” into uncontrolled plant populations. Isolation distances can be quite substantial, especially for species which are considered to be efficient cross pollinators.
3. It is also usually required that all the seeds harvested from the trial should be destroyed (including material from non-transgenic controls), and the site of the trial monitored for several years with particular reference, for example, to the development of offspring from buried volunteer seed.
4. As a result of these special requirements, it is usually impossible to compare results from experiments with transgenic plants directly with results obtained from more traditional trials. One way to overcome this problem is to include all the non-engineered plant lines in the “transgenic” trials, and to treat the whole trial as a transgenic one.

Although this is perfectly feasible for some traits, it creates impossible workloads if undertaken in a conventional breeding trial, where a breeder may well want to compare several tens of lines, only a few of which are engineered.

An additional effect is that it is very difficult to comply with the regulatory requirements for companies or institutions working with engineered lines of their main crop. This apparent paradox is best explained with a practical example. Consider a company specialised in sugar-beet breeding, which has developed transformation technology for its crop.

Field trials with the transgenic sugar-beet would be quite difficult to conduct at the main experimental stations of that company, since these stations will have a considerable number of other sugar-beet trials on those sites, and would therefore usually not meet isolation requirements. This is why so many field releases of transgenic plants have been carried out in environments which are somewhat unusual for the crop under investigation.

A final restriction on work with engineered plants versus traditional material is that regulations differ between countries. Procedures differ even more, and criteria for approval or rejection of an application are sometimes contradictory. This poses serious constraints to multi-site agronomic trials undertaken in several countries simultaneously-a very common type of trial in most breeding programmes with major crops in Europe.

Field Tests in Project Designs:

From the above, it can be deduced that any project in plant biotechnology which includes the prospect of field trials should be designed from day one with the restrictions on these trials in mind. To explain why this is so, it is again best to consider a number of practical examples, and to define which early choices in the project have an impact on the acceptability of the resulting plants for field introduction.

Vector Construction:

The DNA sequence that carries the desired gene into the plant is often a quite complex construct. Apart from the target gene and its promoter and termination sequence (both of which, like the gene itself, can originate from very different donor organisms), the construct usually contains a gene coding for a selectable marker (with its own promoter) operating during the in vitro culture phase of the project.

Often a different marker gene is used to follow the transgenic line in the greenhouse and in the field. If transformation is achieved with the use of Agrobacterium tumefaciens (the most common method) all the pieces of DNA mentioned above are embedded in two border sequences, originating from the Agrobacterium genome.

It is therefore quite common that the DNA inserted in the host plant contains sequences from three to seven totally different organisms. If one of these sequences originates from an organism which is classified as a human pathogen, this can lead to delays in the approval of the recombinant plant for field release, or even to refusal of the application.

Another aspect of vector construction which can create serious problems is the presence of “junk DNA”. This is a rather unfortunate name for non-coding sequences of DNA, usually leftover pieces of sequences flanking one or more of the genes of interest in the donor organism.

Often, scientists do not go to the trouble of eliminating all the non-essential DNA in the early stages of vector construction, partly because these flanking sequences often provide useful sites for linking different genes together.

Another type of junk DNA is a total DNA preparation from a commercial source (e.g. calf thymus DNA) which is used to protect the DNA of interest during direct transformation experiments. Some of this DNA may become inserted in the genome of the host plant. One of the recent evolutions in the application of regulations on field releases has been the insistence on “clean vectors”, containing only DNA of which the function is well understood.

The time lag between the start of vector construction and the first field trial of the transgenic plant is usually two or more years. The cost of starting with vectors unacceptable for field trials is therefore very high, in terms of time and effort wasted. It is important that information on conditions for field releases is widely distributed in the research community.

Most industrial corporations entering a research area are aware of these restrictions, or have procedures in place whereby their scientists check up, at the start of a project, on the regulatory framework in which they are asked to operate. The same is not true for most academic institutions, or for many bodies funding biotechnology research.

Choice of Host Plant:

The selection of the right recipient for the genes to be engineered is crucial in determining the chances of success of later field evaluation of the transgenic plants. Until recently, transformation methods were not sufficiently advanced to allow the scientist a wide choice of host species and/or varieties.

Most of the pioneering work in plant genetic engineering was carried out with tobacco, simply because it was the only species for which the actual production of transgenic plants was not a limiting factor in the research projects.

Somewhat later, scientists started using certain tomato and potato lines as alternative “model species”. When those early projects reached the stage where plants were ready for testing under open field conditions, it became clear that a lot of the work had been carried out with varieties of these plants that were not suitable for field evaluation.

Tobacco as a species actually turned out to be an unfortunate choice, because it was difficult to grow to maturity in the climate of the regions where much early genetic engineering work was done. This forced scientists to make complex arrangements to conduct field trials sometimes thousands of kilometers away from their laboratories.

A problem of a different nature arose with early potato work, much of which was done with a very old variety called Berolina. The only reason for choosing this variety was that the transformation results were quite good. Unfortunately, it turned out to be a variety that was very difficult to transfer from the test tube to the field. Therefore it was impossible to test the performance of the introduced genes (in this case insect resistance) in the field.

Another much used potato variety has none of these problems, but flowers profusely. Since many of the early field experiments were approved on condition that the plants would not be allowed to flower (to avoid spreading of the recombinant genes), trial fields had to be patrolled daily and flower buds removed during the entire flowering period.

Troublesome as these early problems were at the time, they served the useful function of stressing the need to consider very carefully the choice of recipient plant in any new genetic engineering project. Today, it is a priority in most applied projects to establish sufficiently reliable transformation procedures for elite breeding material of the project’s target crop.

This is often a major factor in hindering research. Research teams have repeatedly found, to their distress, that solving transformation for one particular line of a given species offers no guarantee that this will be automatically applicable to other varieties. Nevertheless, the list of plant species of major scientific and/or economic interest for which transformation is now routinely possible using elite germplasm is growing rapidly (Table 1).

Table 1: Routinely Engineered Crops:

Data Collection on Newly Produced Plants:

Transformation experiments, and the plants produced, are constantly monitored for parameters indicating success of the experiments. One transformation experiment will usually generate several (sometimes hundreds) of independent transgenic plants. The main goal of the monitoring process is to find the line(s) that gave the best expression of the desired trait in the absence of any side effects. Many of these observations are also used in the preparation of requests for field releases. This early work is done in growth cabinets or, more usually, in a greenhouse.

Expression of the Target Gene:

The primary interest of the scientist is, of course, to see if the transformed plants express the new traits. This can be done biochemically, by detection of the protein encoded by the engineered gene, or by a measure of its activity. If the new gene also gives the plant a recognisable new phenotype (e.g. insect resistance), bioassays will also be performed.

In the case of insect resistance these assays will include target and non-target insect species. It is quite common to find differences in expression levels of a factor hundred between independent trans-formants. This is due to the effect on expression of the place v here the gene is inserted in the plant genome.

Expression of the marker Genes:

The expression of the marker (or reporter) genes can be very important for later genetic work on the transgenic lines, especially if the target trait of the project is not readily observable (e.g. improved protein composition of seeds). The most widely used marker genes are the resistance genes to the antibiotic Kanamycin and to the herbicides phosphinothricin and sulfonylurea, and the gene coding for the enzyme glucuronidase.

Gene Copy Number:

In most cases it is preferable to obtain the desired effect with only a single copy of the new gene, since this greatly facilitates the subsequent genetic work of transferring the trait in other germplasm by conventional breeding techniques.

Plants are usually screened for copy number by Southern blot analysis on the original trans-formants, followed by a segregation study on the first generation progeny. In those cases where no single copy plants with all the desirable traits are found, multiple copy plants are separated into single copy plants by out crossing.

Plant Morphology:

All plants are carefully observed for visible malformations. These can result from mutations induced by the in vitro growth conditions of the plants, or from the fact that the engineered genes have by chance been inserted in a locus of the plant genome which is important for normal plant development. These observations are usually continued into the second greenhouse-grown generation.

Up-scaling of Material for Field Testing:

This is a critical part of the greenhouse work. Field tests usually require substantial amounts of seed. More importantly, for most crops the quality of the seed produced in the greenhouse is not as uniform as field- produced seed. For some crops, even the best seed batches produced in the greenhouse will perform less than elite field-grown seed in a yield trial. To make the best of the first field trials in a project, it is essential that utmost care is taken to use uniform seed from the onset of the work.

Execution of the Field Trial:

Pre-release activities:

The work on a field trial starts about one year before anything is planted, with the prospection of possible sites. This includes checking whether the expected isolation requirements can be met. It may also include preliminary experiments with control plants to verify that the site is suitable for study of the target gene (e.g. presence of insect pests), especially if the trial has to be conducted in a place where the crop is not usually cultivated.

Trial approval:

Most countries require about three months to process an application file. In practice, and especially if a previously untested gene or crop is proposed, it is better to introduce the file five to six months before the expected field planting date, to allow for additional questions from the authorities.

Execution of the Field Experiment:

The central goal of most field experiments is the continued observation of the new traits of the engineered plants. Methods used, and trial layout itself, will be determined by this. In experiments where the main goal is to compare yield, a conventional statistical design will be used. In disease resistance trials it may be necessary to provide for good inoculation conditions for the disease. There is a large body of experience with this type of trial in plant pathology.

One of the specific features of early trials with engineered plants is the emphasis on further screening of different transformed lines for conformity to the non- transformed control. Although most off-type plants will have been eliminated in the greenhouse stage, subtle side effects of the transformation procedure or of the introduced genes can only be visualised in a replicated growth and yield trial in the field, where the full complexity of a variable environment influences the development of the plants, and where plot size and number is sufficient to allow the often very small effects on growth and/or yield parameters to become statistically measurable. Another aspect of the observations in the field is related to the requirements for monitoring put forward by the authorities in the approval.

The requirements depend on the crop and trait under investigation. Monitoring extends to the period after termination of the experiment, and to the surrounding area of the trial.

Termination activities:

After harvesting the trial, the plant material in the field has to be disposed of. It should be kept in mind that in many trials the “harvest” consists of observations and measurements in different stages of development, and many trials are terminated well before the plants have grown to maturity or set seed.

Depending on the developmental stage of the plants at termination, different effective methods for destruction are used. If the plants are still green and growing actively, by far the most effective destruction method is by spraying them with a systemic herbicide, which ensures that the roots are destroyed along with the above-ground parts. If the crop is mature and dry at harvest, it is often burned, or rotovated into the soil.

Testing insect resistance:

The first field experiments done by Plant Genetic Systems Ltd. (PGS) were on tobacco plants which had been engineered to express a protein of bacterial origin (Bacillus thuringiensis) which kills the caterpillars and other larvae of certain major tobacco insect pests.

However, the trials had to be done in the USA (North Carolina) because these pests are not significant in Europe, and therefore no meaningful data on field resistance could be obtained in Europe. The most remarkable result of these trials was that insect control was much better in the field than in previous growth chamber tests.

This quite unexpected observation was studied further by scientists of North Carolina State University. They discovered that the insect resistant plants, which had not been treated with chemical insecticides, carried a much larger population of predator insects and parasites of the pest than control plots that were maintained insect free with conventional insecticides.

Testing promoter activity:

One of the most important qualities demanded from any genetic system before it can be considered for commercialisation is stability of performance under different field conditions. In genetically engineered crops, two conditions can cause failure to perform: the instability of the gene product (enzyme), or erratic expression of the gene.

The latter is caused by the sensitivity of the promoter to changes in environmental conditions. It is possible to obtain some preliminary information on the stability of enzymes by extrapolating results from in vitro tests (although these should be verified in vivo), but the stability of promoter activity can only be tested reliably in vivo, preferably under the highly variable conditions found in the field.

In two separate field tests, on tobacco and alfalfa, it was demonstrated that two promoters, both of which controlled expression of a herbicide resistance gene, and both of which gave adequate resistance in greenhouse tests, behaved very differently under field conditions, one promoter giving good resistance under all conditions, while the other gave much lower levels of resistance in the field than in the greenhouse.

Predicting behavior of genes and promoters:

One of the most appealing qualities of genetic engineering is that, in principle, the same gene will work in the same way in any organism, provided that the signals controlling its expression function equally well in different receptor species.

This is why much of the early transformation work on monocotyledonous crops (at present mainly maize) is again using the old and very thoroughly studied gene coding for antibiotic resistance and/or herbicide resistance to evaluate the performance of promoters specifically isolated for work in monocotyledonous species, and to find out if the “old” dicotyledonous and non-plant promoters can be relied on in these new crops. The answer to this last question is important: it could mean years of work to find suitable analogous promoters to the trusted workhorses developed in dicotyledonous plants.

Now that greenhouse evaluation has been completed, the planting season will deliver the critical data to finalise the analysis, in the form of a series of maize trials in Europe and the USA.

Multi-Site Yield Component Trials:

Yield is notoriously difficult to measure, because it is an aggregate index of all the influences of the genotype, the environmental conditions and the interactions between these two on the development of plant communities. To separate all these influences, and to single out the effects of differences in genetic, requires testing over a range of environments.

Few such multi-site trials have been attempted with transgenic plants, partly because of lack of suitable transformed plant varieties, and partly because of the complexity of undertaking replicated trials in different countries, which impose different confinement and/or monitoring requirements on the replicas, thereby defeating the basic feature of the trial.

An example of such a trial was a two-year, multi-site study of rapeseed expressing an engineered seed storage protein. The material was tested in 1989 in Belgium, Sweden and Canada, and in 1990 again in these three countries and in the UK.

The interpretation of the results was complicated by the fact that the variety originally chosen for the transformation was genetically heterogeneous, and therefore gave problems with the choice of suitable control lines for performance evaluation. One of the positive results of this project was that in this case it had been possible to obtain reasonably similar conditions for the release in the different countries.

Testing engineered male sterility genes:

The use of conventional male sterile carrier plants for field testing of genetically engineered traits has often been proposed as a good method for eliminating the risk of gene spread through pollen. Recently, male sterility has also been obtained in plants by recombinant DNA technology. The evaluation of this material requires even more multi- site testing than yield testing, since the first requirement of such a system is stability under all the environmental conditions where it may eventually be used.

Field work was begun with tobacco in 1989, rapeseed in 1990 and chicory in 1991. It is anticipated that the present range of trial countries for rapeseed testing will be further expanded, as trial results accumulate, to include the full range of climatic conditions where the trait may be used in rapeseed breeding.

Field Trials: A Canadian Experience:

Canada has a very large agri-food sector. The 67 million hectares of improved farmland in Canada are divided into 280,000 farms, 98% of which are family-operated production units. The majority of Canada’s major field crops have been introduced during the past 200 years (wheat, alfalfa, barley, rapeseed, canola, and soybean). Most of Canada’s crops have undergone significant genetic modification through classical plant breeding during the past 100 years.

Due in part to Canada’s excellent history in variety improvement (yield improvements, quality improvements, and improvements in pest/stress tolerance), producers are not only receptive to genetic modifications but, in many instances, financially support, through producer associations, the research, development and plant breeding efforts associated with genetic improvement of major crop species.

The evolution of recombinant DNA technologies since the mid-1970s, and the emergence of plant transformation technologies in the 1980s, are providing opportunities in Canada for crop improvement through a combination of genetic engineering and classical plant breeding. These opportunities are currently being exploited by a number of public institutions as well as the private sector.

Although Canada had significant experience with plant cell and tissue culture, there was relatively little activity in plant genetic engineering prior to the breakthrough by Jeff Schell, Marc van Montagu and Mary dell Chilton in developing and utilising disarmed Agrobacterium tumefactions in the early 1980s. At about the same time, the National Research Council’s Prairie Regional Laboratory at Saskatoon was transformed into the Plant Biotechnology Institute.

Numerous researchers in public institutions and a few private companies initiated transgenic plant research in Canada. Many public sector faculty positions in plant biology or plant science departments were filled with recruitment trained abroad in molecular biology.

Many faculty members also used sabbatical leaves to retool in plant genetic engineering technologies during the mid- 1980s. Several small private sector plant biotechnology companies emerged during the early and mid-1980s in Canada, recruiting molecular biologists from the USA and Europe.

Canadian plant scientists worked in networks (both domestic and international) to access technologies and transgenic germplasm for continuing development and utilisation. By the mid-1980s, many public and private sector laboratories were handling transgenic plant material, associated transformation vectors, and genes of potential relevance to Canada’s field crop sector.

Field Evaluations:

By 1987, it was clear that transgenic plants under evaluation in several laboratories in Canada would eventually need evaluation in simulated production environments under field conditions. In that year Canada’s federal Ministry of State for Science and Technology commissioned a background report on regulatory policy options for biotechnology. In the same year. Agriculture Canada reviewed international regulations for genetically engineered organisms and drafted a regulatory process for products of plant biotechnology.

Agriculture Canada, through consultation with the federal Department of the Environment and the Department of Health and Welfare, and through participation in a series of regulatory workshops and consultation with a large number of Canadian public and private sector plant geneticists and breeders, developed a preliminary application procedure in early 1988 for controlled field testing of transgenic plants. The first field evaluation of transgenic plants was conducted later that year by Agriculture Canada. Since then, the evaluation of transgenic plants in Canada has grown steadily and dramatically.

Agriculture Canada reviewed more than 100 applications for field evaluations of transgenic plants in 1991, 302 in 1992, 503 in 1993 and 848 during the first half of 1994 (up to 16th May). The University of Guelph initiated field evaluations of transgenic canola in 1989 and transgenic alfalfa in 1990.

To date, transgenic canola and alfalfa families that have been field-evaluated include a variety of alien gene constructs involving DNA from other plant species, microbial species, fungal species and the coat protein of cauliflower mosaic virus.

Characteristics associated with these “transgenes”‘ include antibiotic resistance, herbicide tolerance, modified amino acid composition, male sterility, male fertility restoration, and tolerance to abiotic (physical) and biotic stresses. In all cases, field evaluations of transgenic plants have followed application to, and approval from, Agriculture Canada.

Reviews of individual applications have taken from two to four months and have required fairly specific information on the nature of the genetic modifications of transgenes, the experimental procedures that will be employed for the field evaluation, a description (physical and biological) of the field trial locations, and a description of post trial procedures to prevent entry of transgenic plant material into the food chain or agricultural production ecosystem.

Local farmers and the media appeared satisfied that the research was necessary, and in the interests of the Canadian agri-food system generally. Although no subsequent public discussions have been held, trialspecific information is released publicly each year. To date, no trial-specific concerns have been directed to the University as a result of these information releases.

In spite of the regulatory framework developed by Agriculture Canada, at least one environmental activist group has criticised the government for permitting field evaluation of transgenic plants. Both the government and the research community (public and private sector) have been criticised for secrecy and irresponsibility associated with field evaluations of transgenic crops. Criticism has been of a general rather than specific nature, intended more to elicit public fear or outrage rather than to address specific concerns regarding human health, environmental protection or sustainable agricultural ecosystems.

Monsanto’s Experience in Developing Plant Biotechnology Products:

The Monsanto Agricultural Group has had years of experience in conducting field trials with a number of crops improved through plant biotechnology. The early experience gained from field testing of insect-protected cotton plants in the USA pointed towards the importance of Monsanto’s policy of being open with respect to communication with the public.

Through modern tools of plant breeding and biotechnology, Monsanto has successfully incorporated a gene from Bacillus thuringiensis (B.t.), a naturally occurring soil microorganism, into cotton so that the cotton plant effectively controls the larvae of caterpillar insects. Monsanto researchers worked for many years before they were ready to take this cotton to the field for its first tests in 1991.

During the first year of field testing, 20 small-scale tests covering a total of three hectares were conducted. An essential element in gaining regulatory approval and successfully conducting the tests was open communication with many of the audiences that were interested in this work. These audiences included Monsanto employees, governmental regulators, cooperators, members of the community near the test sites, and members of interest groups.

Openness in many aspects of developments in plant biotechnology has been very important in gaining freedom to operate throughout the earliest stages of work. General public awareness of science and new scientific principles in the USA is not well advanced.

However, there is a predominant view in the public arena that science is neither intrinsically good nor bad, but is dependent on the people who are undertaking scientific research. Given this situation, public confidence can be enhanced when those involved with the work communicate about it. Early discussions with key community leaders, government officials, educators and others have been well received and have proven invaluable.

In the USA, a number of regulatory agencies in addition to the United States Department of Agriculture (USDA) are interested in field studies, depending on the crop and objective of the study. If the new plants contain pesticide traits the Environmental Protection Agency (EPA) will be involved, and varying requirements are necessary depending on the size of the test and the disposition of the crop products. If the crop products will be used for food and/or feed, the Food and Drug Administration (FDA) may be interested.

At the state level, several states have their own requirements which should be considered prior to field testing. Monsanto provided information to all of the appropriate regulatory agencies very early on in the development of policies and in the conduct of specific studies. One of the consequences of early and regular open communication with these organisations has been the establishment of mutually respectful relationships.

Monsanto has found particular value in working closely with its partners and cooperators in the early planning stages of its field trials. Site selection is important as it involves both agronomic criteria and non-agronomic criteria such as convenience to facilities and security. In plant biotechnology research, the design of this year’s trials often depends on the previous year’s test results.

Unfortunately, the time between the collection and analysis of last year’s data and the regulatory deadline for the submission of the paperwork requesting the next round of tests is very short. We have a very narrow time-frame in which to meet the research, planning and administrative requirements. Again, good communication is a key to successful completion of these requirements.

Following test approval, we have a second very narrow time-frame in which to communicate with all of the individuals and organisations that require specific information about the test. We must ensure that all test requirements are met, from shipping, receiving and, storing seed to planting conditions.

Biotechnology and Pioneer Hi-Bred:

Pioneer Hi-Bred International Inc. is one of the world’s largest independent seed companies. Its goal is to deliver improved genetics to the world’s farmers. To achieve this goal, the company has strategically placed research stations around the world, including several in South America. Pioneer Hi-Bred’s major products are hybrid maize, sorghum, soybean, sunflower, canola, alfalfa and wheat.

We expect biotechnology to play a major role in the continued improvement of all these crops, through the technologies of gene mapping and genetic engineering. Pioneer does not see biotechnology as a unique technology producing a unique stream of products, but rather as another tool to be used by its plant breeders in their continuing search for improved germplasm.

Biotechnology in Plant Breeding:

In order to be useful to the plant breeder, germplasm derived through biotechnology must be readily available for use in breeding programmes. The scale of breeding programmes for Pioneer’s major crops dictates that plant breeders make selections in the field, not in the greenhouse, so the ability to plant genetically engineered material in the field is vital to future crop improvement programmes.

The development of new crop varieties is a finely tuned process. Hundreds of thousands of new genetic combinations are tested each year in the USA alone. Maximum use is made of breeding nurseries in locations such as Hawaii, Puerto Rico and Mexico, so that more than one generation of a crop may be grown in a single year.

Potential products are subjected to wide-scale performance testing in areas where they may be sold, including thousands of side by side comparisons in growers’ fields. Finally, there is a rapid scale-up to commercial seed production of those few lines that are finally selected for commercialisation.

If genetically engineered materials cannot be integrated into this system, but are subject to unique barriers, whether raised by technical problems, unnecessarily restrictive regulations, bureaucratic delays or unreasonable costs, then the use of biotechnology in the seed industry will be curtailed, ultimately depriving farmers, processors and consumers of the benefits of crops that use fewer chemical inputs and have improved agronomic performance.

Pioneer supports sound science-based regulatory oversight of biotechnology research and product development, for the public confidence that such regulations provide and the high standards they promote. Clear delineation of the regulatory review process for such products will promote further development of the technology. However, regulations designed to safeguard the environment and human health must also be flexible in order to facilitate the conduct of safe research activities.

Field Testing of Maize:

Pioneer does not feel that biotechnology poses an entirely new set of safety issues. There is a continuum of concern that should be taken into account when establishing a system of regulatory oversight. In some instances, biotechnology derived products will be directly comparable with traditionally derived products.

For example, Pioneer has developed herbicide resistant maize hybrids derived from plants carrying a natural mutation in the acetolactate synthase gene. An equivalent product can be derived through genetically engineering the same mutation into the acetolactate synthase gene and transforming the gene back into maize.

The genetically engineered version does not seem to raise any unique safety concerns, so why should it be subjected to additional regulatory oversight? Regulatory oversight should be based on the risks, if any, that are posed by the product. Plant breeders have always sought to introduce new characteristics into domesticated plants, often employing wide outcrosses to wild relatives to access those characteristics.

Confidence in the safety of the resulting material is based on the breeders degree of familiarity with the germplasm, focusing particularly on any known toxins associated with the plant, and on the extent to which the genetic contribution of the wild species has been diluted by backcrossing.

Typically it might take at least six generations of backcrossing to recover the recurrent parent with the new trait. During this process the genetic contribution of the wild species has been reduced from 50% to less than 1%, representing perhaps 500-1000 genes. By these standards, changes brought about by the genetic engineer, who introduces only two or three genes, are amazingly focused.

Many “new” genes will actually be somewhat familiar, because they are derived from plants with a long history of consumption as food or feed, but which are being moved into different crops. Examples might include the methionine-rich seed storage protein gene from Brazil nuts (Bertholetia excelsa) that Pioneer has introduced into soybeans in order to enhance the levels of essential amino acids; or wheat germ agglutinin, a plant lectin that inhibits the growth of European corn borer, an important pest of maize.

Pioneer sees little inherent risk in field testing such genes in these crops, but agrees that it should proceed with caution until questions about the possible adverse effects of altered patterns of consumption of these proteins (e.g. potential food allergies) or impact on non-target species have been resolved.

The North American Experience:

Pioneer has developed the capacity to genetically engineer most of its key product crops. It has field tested transgenic alfalfa, maize, canola, soybean and sunflower in North America. All these trials have been reviewed by the United States Department of Agriculture (USDA) or by Agriculture Canada, as well as officials from the state or province in which the trials were conducted.

The test protocol also has to be approved by the Institutional Biosafety Committee, a group of Pioneer scientists and managers with experience in biotechnology, agriculture and risk management, and two outside experts in related disciplines. Pioneer has also taken steps to inform local opinion leaders and the press of its activities.

Undoubtedly, much of the success of the field testing programme in the USA is due to the scheme of regulatory oversight, outlined in the 1986 Coordinated Framework for the Regulation of Biotechnology, as implemented by the Biotechnology, Biologies and Environmental Protection unit of USDA’s Animal and Plant Health Inspection Service (APHIS).

By focusing on scientific principles of risk assessment and avoiding judgments based on speculation, they have earned the respect of the regulated community and discouraged the proliferation of local regulations. Local regulations, while often well intentioned, are seldom consistent and frequently duplicative of national regulations. From the applicants standpoint, the advantages of dealing with a single regulatory entity are considerable.

The USDA system is certainly a model worthy of serious consideration by any country seeking to implement regulatory oversight of transgenic crops. However, it is not perfect, and some of the problems will be discussed in the context of Pioneer’s experiences in field testing genetically engineered maize.

Field Testing Genetically Engineered Maize:

Maize represents the prime target for most major seed companies but has proved a demanding technical challenge. Within the last two years. Pioneer and several other companies have finally achieved stable transformation of maize. This has been possible through use of the particle gun to accelerate micro projectiles, bearing recombinant DNA, into plant cells, coupled with tissue culture responsive cell lines. Initially, regulators expressed some concern as to the stability of recombinant genes introduced using this technology.

This resulted in demands to see molecular and genetic evidence, in the form of Southern blots and segregation data, confirming stable integration. Recombinant DNA introduced in this way has not proved unstable and does not replicate independently without a plant origin of replication.

Pioneer does not have the resources to conduct molecular or genetic characterisation of every construct that may be field tested. After all, it may take 100 independently transformed lines to identify one that has adequate expression of the recombinant gene under field conditions.

If molecular and genetic characterisation is required for regulatory purposes, then it should be done on those lines that are finally selected for commercial advancement. The genes that Pioneer has introduced into maize and has (or will shortly) field test include two herbicide resistance marker genes, two marker genes controlling pigmentation in maize, two dominant inhibitors of gene expression both derived from maize genes, three different viral coat protein genes that may confer resistance to important maize viruses, and wheat germ agglutinin.

Pioneer does not consider that any of the genes described represent a high risk, either in terms of enhancing the weedy properties of wild relatives of maize or in terms of human or animal health. However, Pioneer is conscious that some of these genes, and the control sequences used to achieve expression, do not occur in domesticated maize, and care should be taken to limit dissemination of these genes through seed or pollen until their properties have been adequately evaluated.

Maize itself is a highly domesticated crop and there are no feral populations. It would take fundamental changes in the biology of the plant to make it an invasive weed. There are no wild or weedy sexually compatible relatives of maize in North America, so there is no genetic bridge by which recombinant genes could find their way into wild populations.