Jonathan Jones
27 January 1995
Modern plant breeding is firmly based on the principles of genetics, but genetic modification of crop plants commenced long before Mendel discovered the principles of heredity. In the last 12 years, an additional technique for plant modification has been developed, that permits insertion of designed recombinant DNA molecules into plant cells. Here, I assess some of the issues this raises.
First, a definition. I use the term transgenic plant technology (TPT) to refer to the production of plants into which defined new DNA sequences (transgenes) have been introduced. Such transgenes are manipulated in the laboratory so that upon incorporation into plant chromosomes, they express defined new proteins. They usually make a tiny contribution to overall gene expression compared to the 20,000 or so genes that the plant already contains, but under certain conditions the consequences of these transgenes can be very important. Transfer of new genes into plant chromosomes is nothing new; the soil bacterium Agrobacterium tumefaciens is a plant parasite which delivers specific DNA sequences in the course of its life cycle, and most of the transgenic plants produced in laboratories have been made by borrowing the Agrobacterium gene transfer system and modifying it so that only desired transgenes are transfered. Part of the difficulty for TPT enthusiasts is that they and their claims are viewed through the prism of social anxiety about the consequences of other technologies. Despite the claims of nuclear scientists in the 1950s that splitting the atom would make electricity "too cheap to meter", the cost of making nuclear power as safe as it is (sic) has proved to be enormous. When new pharmaceutical and agrochemical compounds were developed and distributed without adequate anticipation of dangerous side effects, public scepticism about new technologies was exacerbated. Small wonder that the public lacks confidence in the latest assurances from scientists. This lack of confidence is particularly galling because many TPT participants, even those in companies, believe that a real value of their work lies in the potential for creating plant genotypes which require less application of agrochemicals for efficient plant production.
The current routine deployment of TPT raises some specific questions.
(1) Does the use of TPT for plant breeding differ in kind from previous techniques to obtain useful genes for crop improvement?
The introduction of genes from other species into crop plants has been a widely used strategy in plant breeding, especially for breeding disease resistance. For instance, genes for virus resistance have been introduced into the tomato (Lycopersicon esculentum) from the wild species L. peruvianum. In so doing, breeders introduced numerous (and uncharacterized) L. peruvianum genes in addition to the virus resistance gene. It seems peculiar (to TPT protagonists) that modern technologies, which enable scientists to control the gene transfer process and introduce specific genes (and no other genes) into a crop plant, should be perceived as potentially more dangerous than such tried and tested but imprecise breeding techniques. The new technology, it is argued, simply increases the species range and the precision of gene transfer. This is a reasonable view, though for some people, the extension of the species range to include all forms of life constitutes a qualitative, not a quantitative change.
There are those who say that "man" should not transgress species barriers. This is not so much an argument as a deeply held belief, which I can respect but do not share. Why is it any worse to transgress a cellular barrier with DNA than to transgress a geographic barrier by, say, growing tomato and potato in Europe, when they originate in Latin America ? If Sir Walter Raleigh had faced these kinds of concerns, he might have decided to leave tomato and potato in the Americas. The import of these plants provoked major changes in European farming, but certainly not an environmental catastrophe.
Undoubtedly, many human activities such as the production of food and other items, transportation, forestry etc affect (mostly adversely from the viewpoint of other species) the whole planet. We have to manage the global ecosystem better, and to deploy technologies with less environmental impact. There is no reverting to some bygone time before we ever changed the environment. TPT should play a crucial role in developing a more ecologically sound agriculture.
(2) Can the public be satisfied with the competence and motivation of the participants?
The desire to do something useful is an important motivation of those involved in this work, though it would be naive to pretend that some participants are not also strongly driven by the desire to do important, "high visibility" science, or to get rich (or both). However, the technology is so easy and so safe that in the event of a TPT-derived plant being produced which contains unintended novel DNA, the chances of it constituting a threat to people or the environment are extremely low. Additionally, such a plant would be unlikely to get beyond the greenhouse, because in the course of its analysis prior to varietal release, the introduced DNA would be thoroughly described and the plant would be discarded from trials if it proved to contain the wrong DNA. In other words, TPT seems at present to be safe and reliable. The people working with it are competent. Even if they weren't, TPT is so safe that the chances of any error leading to a public health hazard are infinitesimal compared with the risks to society and the environment from errors by people in the oil industry, the pharmaceutical industry, the chemical industry and the nuclear industry. It is frequently pointed out that an accidental release of a living organism is different from the release of a chemical, because it can reproduce. The "genie cannot be put back in the bottle". But if it is an engineered plant which is perfectly safe, and easy to eliminate from inappropriate locations with herbicides, so what? An analogy is also drawn between the potential deleterious consequences of the cultivation of transgenic crop plants and the known problems caused by the release of feral plants and animals into new habitats, such as the prickly pear or the cane toad in Australia. But this is a misleading analogy. To grow a plant with one or two extra genes in its complement of 20,000 genes, in an area where it has been grown for centuries, is quite different from the release of a plant into a geographical region where it has never been seen before.
(3) What traits have been or will be confered on plants using TPT, and why?
The most well known example of TPT is the FlavrSavr tomato, in which a gene involved in softening tomato fruit during ripening is switched off. This results in a fresh market tomato that can be picked later, when more flavour has developed, without rotting so much in transit that it is unsaleable on arrival in the store. It also results in an improvement in the quality of processing tomatoes, permitting more solids to be recovered. Thus, instead of 2 genes being added to the 20, 000 plant genes, one is added, and one plant gene is inactivated. The added gene confers antibiotic resistance, and is required to select plants cells carrying the desired DNA. Such antibiotic resistance genes are also not an environmental threat; many or most of them originate from soil bacteria in the first place.
Another major commercial objective of TPT is to produce plants which are no longer sensitive to plant pests. For example, if cotton plants could be produced which are resistant to boll weevils or other insects, spraying of insecticides could be drastically reduced. This is seems to be an obvious good thing, even if (as is likely), pathogens eventually evolve to overcome the resistance, and new transgenes or combinations of transgenes must be regularly introduced for continued pathogen control. Transgenes for control of viral diseases are also well developed, though transgenes for control of bacterial and fungal diseases have lagged behind. However, recent discoveries, in my own laboratory amongst others, have increased optimism about the prospects for TPT approaches to the control of these pathogens too.
Another major objective of TPT has been to engineer herbicide resistance. Readers are excused for experiencing scepticism about the worthiness of this endeavour, and anticipating that herbicide-resistant crop plants will increase herbicide applications and the resulting environmental pollution. However, there are several compounds which (when pure) are of negligible mammalian toxicity, are effective herbicides at very low dose rates, but kill (nearly) all known plants (even the crops, thus limiting their use). Such compounds include sulphonylureas (made by DuPont) and glyphosate (or Roundup, made by Monsanto). If one accepts, however reluctantly, the idea that herbicide applications are an unavoidable part of actually existing agriculture, it is much better to use these herbicides than to use more toxic, higher dose rate compounds such as 2,4 D and 2,4,5 T of Agent Orange notoriety. The new, more potent herbicides cannot be widely used at present (except pre-emergence) unless crop plants are engineered to herbicide resistance. Not surprisingly, much of this engineering effort is taking place in the herbicide companies, since if they can engineer resistance, they can sell more compound. This objective need not generate particular public enthusiasm, since the major economic beneficiaries are the companies, but neither should it generate public opposition, since the outcome will principally be to substitute the use of more dangerous compounds with less dangerous compounds (though there is at least one example of engineering 2,4 D tolerance in cotton).
Other likely applications of TPT include novelties (e.g. new flower colors, such as blue roses), changing the oil or starch composition of crops, and the production of new speciality chemicals and pharmaceuticals in plants. A particularly exciting recent development is the production of biodegradable plastics in plants. As any parent knows, the patter of tiny feet is often accompanied by the swelling of the familial dustbin, and the municipal tip, with large volumes of used disposable nappies. The plastic in these nappies is not biodegradable. It seems likely that the biopolymer polyhydroxybutyrate (PHB), produced in plants, could be incorporated into biodegradable nappies, reducing this waste problem.
(4) Will the promises about the benefits of TPT be fulfilled?
Will the substantial investment of public and private research funds prove to be worthwhile? There will be successes and failures. It remains to be seen how quickly pathogens will overcome engineered resistances, for example. The time scale for success in engineering new oilseed composition may be too long for the patience of investors, though again, recent progress has been spectacular. Overall, the probability of real new use value in society is very high, and in my view outweighs any hypothetical causes for concern.
(5) What will be the consequences of TPT for the mode of production in agriculture, for the food industry, or for other consumers of plant products?
Predicting the impact of TPT on the mode of production and on agricultural markets is complex. Seed prices may go up, but chemical applications may go down. It is unlikely that TPT-modified crops per se will dramatically accelerate the changes in agriculture which lead to small farms going out of business. TPT modification of crops could extend the growing region if the crop became more profitable (e.g. oil seed rape would be more widely grown in the US if modified to a higher value oil composition). TPT objectives and consequences are strongly influenced by the state of existing markets. If sugar beet could be engineered to produce something more valuable than sugar (e.g. speciality pharmaceuticals), then TPT would lead to the retention of a crop which might otherwise gradually disappear if EEC subsidies were phased out. If speciality chemicals and pharmaceutical products came to be produced on a large scale in plants (a real possibility), less land would be available for food production. Land which farmers are paid not to farm would come back into production and food prices would probably rise. On the other hand, there would perhaps be fewer chemical plants or oil spills if more chemical feedstocks derived from plant biomass. It is impossible in a paragraph to do justice to the range of possible consequences, and this area merits careful future attention.
(6) Who will own the products of this technology and how will this ownership be protected?
In my view, the most serious controversy surrounding TPT concerns the ownership of plant genotypes. Traditionally, in both Europe and (to a lesser extent) the US, new varieties developed by breeders are protected by Plant Varieties Rights Protection (PVRP). Other seed merchants can sell the variety but the breeder can charge a royalty on that sale. Most importantly, breeders can use varieties developed by their competitors as a source of germplasm in their own breeding programs. However, Molecular Genetics Inc. of Minnesota was awarded patent protection on an entire plant which arose through selection for herbicide resistance in cell culture. This is being seen as a precedent for permitting companies which develop new (patented) transgenes to also patent the resulting transformed plants. A consequence of this would be that even those useful genes in such a transgenic variety which had not been introduced by TPT would be unavailable to other breeding companies. The most cynical interpretation of this would be that TPT is thus a Trojan Horse by which major chemical companies will more easily compete with existing seed companies in a seed market. This market will in the long term replace chemicals as the route by which new crop protection technology enters agricultural practice. I believe it to be essential that patent protection does not extend beyond specific transgenes, and I condemn the decision of the US patent office to award a claim to "all transgenic cotton" to the American company Agracetus. AgraCetus' competitors are also taking a dim view of this decision which is likely to be subject to lawsuits.
The implementation of a system based on transgene ownership will be complex. Every crop variety will need to declare its transgene composition in the varietal description. Genetics companies will doubtless test major varieties which are declared to lack their transgene to ensure that competitors are not trying to evade royalty payments. One can imagine in 10-20 years that different crop varieties will contain dozens of different transgenes, and keeping track of who is owed how much royalty for which transgene will keep armies of lawyers and accountants fully occupied.
Should items derived from TPT-modified plants be labelled as such in supermarkets? When a commodity such as tomato paste or cooking oil is involved, this seems unrealistic because raw materials are carried over during processing. Processors cannot be expected to clean out their factory every time there is a switch from processing a TPT-modified raw material to a non-TPT-modified raw material. However, advertisers may well seek to make processed products more attractive with slogans such as "Brought to you better and cheaper through TPT". Companies that produce brand name TPT-modified produce such as tomatoes may well insist on labelling to preserve customer recognition. TPT-modified tomatoes in stores in the US sell out as soon as they are delivered, despite a considerable price premium.
(7) Could transgenic plants constitute a public health hazard?
As mentioned earlier, it is unlikely that commercial applications of TPT could create a public health hazard. The primary effect of TPT is to produce plants which make proteins that they didn't make before. Most proteins are broken down to non-toxic components in the stomach. However there are three unlikely but imaginable routes by which toxic consequences could arise due to TPT. (i) The protein could be toxic (such as some proteinase inhibitors, which may be introduced for insecticidal purposes). The use of such insecticidal proteins should be restricted to plants which are not eaten by humans e.g. cotton (ii) The protein could convert a non-toxic plant compound into a toxic one. This is unlikely, but out of concern for unanticipated toxicities it is clear that early candidates for release will go through extended feeding trials. (iii) The protein could convert a sprayed compound into a toxic compound. For example, some herbicide resistances work by enzymatically converting a herbicide into a compound which is non-toxic for the plant. Would this new compound be non-toxic to mammals? What about plant enzymes that could act on the altered compound? This is an expensive area for the TP technologists. Because of these considerations, before any TPT-produced vegetable hits the supermarkets, it will probably have been through more extensive testing than any vegetable in history. In a litigious society, the financial consequences of letting out a dangerous new edible product are appalling, especially to the small companies who are seeking to turn their scientific expertise into revenues. On the other hand, the costs of testing such products are also very high and could keep all but the largest companies out of the business. This would be regrettable and probably unnecessary, since most of the introduced genes will encode proteins which are known or can easily be shown to have or cause no mammalian toxicity. Recently in the US, the route to FDA and EPA approval the route to registration has become easier. I think this is to be welcomed.
What is needed for useful discussion of this topic is the capacity to think sensibly about risk assessment. What, if any, level of risk is acceptable? Society has to answer this question in relation to all new technologies , not just TPT, bearing in mind not only speculative hazards but also the potential benefits and the extra costs of those benefits if the technology is held back by the high cost of testing. In order to assess this, it essential for those who profess to speak for the public interest to be well informed about what recent advances in our understanding of genetics really mean.
(8) Can TPT be used be used for anything other than good works?
It is, regrettably, possible to imagine how TPT could be used deliberately for bad ends. Plants (e.g. lettuce) could be engineered to express proteins that were toxic to humans (e.g. ricin). This could find use as an assassination tool by secret services (who would suspect the lettuce?). This ugly possibility behoves all responsible scientists to undertake not to conduct such research for the military. It is also conceivable that in several decades time, making such transgenic plants would be in the capability of the kind of people who put cyanide in grapes, or indeed, bombs on aeroplanes. But it makes no more sense to oppose TPT technology because it can be put to bad ends than it does to oppose the glass industry because lunatics can put glass fragments in baby food.
This short review barely scratches the surface of a major new topic. I have also avoided the area of veterinary and animal husbandry applications of biotechnology and transgenic animal technology (TAT). The controversies over bovine somatotropin (BST) and the production of pharmaceutical proteins in the milk of transgenic animals require careful thought. However, in general I believe it is time to move on from fundamentalist arguments about whether or not TPT and TAT should be done at all. It is time instead to consider on a case by case basis the specific economic, safety, environmental or ethical animal husbandry issues raised by each proposed implementation of the technology. Particularly frustrating to scientists are the differing stages of the debate in different countries. For example, in Germany, there was vigorous public opposition to a field experiment involving petunias modified to express a gene that provides a slight difference in flower color, whereas in Spain, the UK and especially the US, many field trials of plants modified by TPT have taken place.
As hinted at the beginning, I believe that a major reason for public scepticism has been the history of flawed public pronouncements about safety and economy by the nuclear and chemical industries. This led in some countries (especially Germany) to a complete impasse in the debate on TPT. Opposition groups have taken the view that noone involved in TPT can be trusted to comment objectively on its safety, since these are the people who want to do the experiments. On the other hand, they have noone else to turn to for information. A respected German scientist who uses TPT, and who is also a long term antinuclear campaigner, has been accused of undertaking antinuclear work for many years purely to develop public credibility so that when TPT came along, he could hasten its adoption. Such accusations do the public interest a disservice, but they indicate the depths to which public confidence in scientists has sunk. Scientists need to think carefully about why this trust has been lost, and to behave in ways that earn and deserve such trust, before they can legitimately complain about the "loony fringe" which opposes their work.
Professor J. D. G. Jones heads a research group at the Sainsbury Laboratory based at the John Innes Centre, Norwich, UK, working on on the genetics and structural and functional analysis of plant disease resistance genes and their manipulation, with particulare reference to Arabidopsis (thale cress) and tomato. Email: Jonathan.Jones@bbsrc.ac.uk
Prof. Jones' web site can be accessed via the home page of The Sainsbury Laboratory at the John Innes Centre, Norwich UK