OECD Conference on the Scientific and Health Aspects of Genetically Modified Foods MEMORANDUM by Arpad Janos PUSZTAI 6 Ashley Park North Aberdeen AB10 6SF 1. What do we know about the current stage of scientific developments in the genetic modification of foods; what new GM techniques are on the horizon? The present method of gene transfer which enables scientists to transform any plant using virus and antibiotic resistance genes and which is now the dominant technique for the creation of GM crops, has had a negative influence on finding new ways for genetic manipulation. As a result, most GM crops in existence today are the products of the same and rather imprecise technology that may harm both consumers and the environment. The use of naked viral DNA promoters which are known to be hotspots of recombination with host DNA, the inclusion of antibiotic resistance genes in the gene- transfer construct and the unpredictability of both the site of insertion and its consequences for the plant genome makes this method unacceptable. In addition, present day GM-crops designed for increased pest resistance are not sufficiently selective and specific for their major pests and, by inflicting damage to beneficial insects, they destroy the natural balance between pests and useful organisms. To have a future, we shall have to concentrate on exploring new ways for gene transfer or modification. Thus, viral promoters will have to be replaced by genetic material from plants whose efficiency will have to be improved. Other new and presently unknown GM techniques, such as the recently proposed chimeraplasty, will have to be explored. There will also have to be more fundamental research conducted into gene organization and interactions to be able to predict and control gene silencing and reinforcement of gene action in the plant genome to improve our understanding of gene transfer and to make it less of a hit-and- miss affair. 2. What can be said about the potential effects on health of GM-foods? Are health risks more likely with GM compared to conventional foods? The often-heard statement that GM crops are just another cultivar is simply untrue. No viral, bacterial or mammalian DNA found in present day GM crops resulting from the process of genetic transformation could have been introduced into the plant genome by natural means or traditional cross-breeding. GM-food therefore contains foreign genes and their products that had never been eaten before and whose effects on health and metabolism are unknown, unpredictable and untested. Although most nutritional journals are full of papers of animal feeding studies in which the nutritional value and potential harmful effects of plant based conventional feedstuffs are evaluated, only three such studies with GM-crops have been published in peer-reviewed science journals (1-3). The situation with GM-food is worse because, according to Sir Robert May, they have never been tested with human volunteers before their release into the human food chain even though about 60-70% of our processed food is reckoned to contain GM soya and maize. It is also expected that with the likely prospect of the inclusion of more GM-crops into the human diet in future, such as unprocessed/uncooked greens, vegetables, fruits, etc, the potentially harmful effects of foreign DNA and gene products on human/animal health will be substantially increased, particularly because of "tradition" these crops are accepted as a matter of course and without proper testing. However, in this new age of food technology in which crops and other basic human dietary ingredients are being changed, manipulated and developed at breakneck speed, reliance on traditional values of food alone, particularly of the GM-variety, is fraught with unpredictable and potentially dangerous consequences. 3. What is the science basis for the regulatory framework involved? How risk is assessed within that framework? GM-foodstuffs are presently accepted on the basis of their "substantial equivalence" to their non-GM counterparts and it is not required by the regulatory authorities that the results of biological risk assessment or nutritional/physiological studies carried out with the GM-crops should be published in peer-reviewed journals. This is regarded as wholly unacceptable by most people. The regulatory framework almost exclusively relies on the unpublished results of "in house" work of the biotech companies contained in their submission to the regulatory authorities for scrutiny when they wish to introduce a new GM crop. Clearly, even if these are scientifically valuable, they fall down on the public's (and other scientists') desire of transparency. Moreover, when the regulators wish to confirm or reject any of the results in the submissions, their hands are tied because they cannot commission new work and their judgement on risk assessment is therefore based alone on information received from the companies. Although they can ask for more work to be done when, in their opinion, this is needed, this may not be the best solution, nor is it conducive to the transparency demanded by the public since the information submitted by the companies to the regulators is usually confidential. It also appears that the lines of communication between advisory committees and active scientists working in GM- related fields may not be adequate and likely to be dependent on personal knowledge of what is going on in the labs and fields and by whom. In a European context the situation is even more acute. In fact, if nothing else was demonstrated by our case with the GM- potato work, it showed that once I was gagged, the results of our work could not reach the Advisory Committees. There is a clear need for a route through which, when needed, results could more directly be passed on to the Committees. As the forte of most of the members is scientific administration, the Committees should not only be strengthened by the presence of consumer and environmental pressure groups but also by the appointment of active scientists of different disciplines. It is unlikely that due to the slow publication of scientific papers, without a network of contacts with workers at Institutes active on the field, members of the Committees will be able to keep up with scientific progress. Whether it is liked or not, they will have to be more proactive/interactive with scientists and get information about unpublished work in confidence as early as possible, perhaps through an overarching body of regulators or the FSA to be set up by the Government, such as recommended by the House of Commons Science and Technology Committee. However, it has also to be able to fund research of its own, be independent and, most of all, its work must be fully transparent. This would be greatly helped if, before permission for the release of a GM-crop is granted by the regulators, the information on which their decision is based, is fully published on the internet. Biological testing 1. The whole science basis of the regulatory framework and risk assessment will have to be fundamentally re-thought. This framework that almost exclusively relies on the concept of substantial equivalence needs to be put on more rational and scientific foundations. The apparent lack of significant differences in raw compositional data between transformed and non-transformed lines is singularly unhelpful as, depending on growth conditions, composition can vary within wide ranges regardless whether parent or GM lines are taken. As after transformation many selection and potential back-crossing steps are usually made, the strict scientific basis of comparing the parent with its GM- lines is no longer there. For meaningful comparisons therefore the parent and transformed lines must be put through the same conditions and grown under identical and strictly controlled conditions, preferably right from the beginning and treated and harvested the same way. To establish true average compositional values and their range of variability, several of the individual plants of the parent and different GM lines respectively should be analyzed. It would be desirable if, in addition to comparing macro constituents such as proteins, starch, lipids, etc, of the parent and GM-lines, their contents of known biologically active components could also be compared. Unfortunately, possible new components arising out of non-intentional compositional changes would in most instances escape detection. All the same, a thorough analysis of the GM-material and its comparison to its non-GM counterpart is an essential first step in the testing process before its release. 2. Reliance on in vitro tests cannot be accepted, particularly when in vivo tests are available. Thus, the assertion that a GM-product is degraded by acid or pepsin or other proteases when the assay is done in the test tube is not acceptable. This has to be established in the stomach and intestines in in vivo experiments because it is common knowledge that many plant proteins/lectins which are degraded in vitro are fully stable in the gut in vivo. In most instances the survival of GM-material in the gut can easily be demonstrated in vivo. Thus, the amounts of GM-products can be estimated by immuno- ELISA analysis or other more conventional chemical techniques in gut washings of rats (or other animals) gavaged with the GM-material. Additionally with GM-lectins, including the Bt-toxin (Bacillus thuringiensis toxin) the presence/absence and/or epithelial binding of the GM-material should also be demonstrated by immunohistology. Similar considerations should apply to the survival or degradation in the digesta of foreign DNA, including the gene construct and its parts such as the promoter, the antibiotic resistance marker gene, etc. 3. It is not sufficient as done presently that the stability, biological, immunological hormonal properties and allergenicity are established with the recombinant material from E. coli rather than with the GM-product isolated from the GM-crop. It is well-known that post-translational modification, including glycosylation, amidation and proteolytic processing, etc is done differently in prokaryotes and eukaryotes. These processes are usually different even in different eukaryotes. As all these have an important bearing on the biological behaviour of the gene-product, for its testing it must directly be isolated from the GM-crop and characterised and possibly be identified with the product of the same gene in its natural plant. 4. There is now wide agreement that animal tests must be incorporated into the GM regulatory process even though most peoples' views diverge about the precise details of what is to be included in the tests. However, some general principles and methods can be "borrowed" from feedstuff evaluation studies. This is where the whole process of the biological testing of GM-food had gone, in my opinion, in the wrong direction with the introduction of the concept of GM-food "toxicity". We have long ago abandoned of the use of genes coding for substances which are poisonous for mammals and the likelihood of creating lethal toxins in GM-plants is low. In any case this could be easily demonstrated by feeding studies with animals. We need to look for more subtle effects of GM-foods on metabolism, organ development, effects on the immune and endocrine systems and the gut flora, particularly in young animals, which determine whether the young will turn into proper healthy adults. These are routinely done in animal nutrition and most of the methods are easily adaptable to GM-food testing. It is the best to start the feeding trials with young rapidly growing rats or other rodents before progressing to those with farm animals. Appropriate iso-proteinic and iso- energetic diets will have to be formulated in such a way that most of the dietary protein is preferably derived from the plant material to be tested. The following diets need to be tested as a minimum: GM-diet, parent-line diet and a diet in which the parent line is supplemented with the isolated gene product at the same level as it is expressed in the GM-line. Additionally, another diet group may also be useful in which the parent line is supplemented with the gene product isolated from its original source but again at the same concentration as it is expressed in the GM-line. The animals to be fed these diets are selected into groups of animals of similar weight (at least five animals per group) which are separately housed to allow the collection of individual urine and faecal samples for the determination of Net Protein Utilization (NPU), Nitrogen Balance, Feed Utilization Ratios, etc. First, short-term experiments are performed under strictly controlled conditions of pair-feeding. Then, the feeding tests could be extended. The animals should be blood-sampled before, during and after the completion of the experiments and the sera used for immune studies (lymphocyte proliferation assay, Elispot, etc) and the estimation of hormones (insulin, CCK, etc) and other blood constituents. During the experiment the animals are weighed daily and any abnormalities observed. At the end of the experiments the animals are humanely killed under anaesthesia, preferably 2 h after the last feed, fully dissected, the gut rinsed out and the contents saved for possible further studies (analysis of food components, gut enzymes, GM-products, DNA, etc), suitable sections taken for histology, wet organ weights recorded and finally the organs and carcasses freeze-dried and weighed again and subjected to compositional analyses if needed. Results are then subjected to suitable statistical analyses, such as ANOVA, paired comparisons by student's t-test (or Tukey's test) and most importantly to multivariate analysis to identify possible contributions by the transgene expression and the construct to the significance of any differences. It is clear that if in these feeding tests any of the above variables show significant differences between the animals fed GM- and parent-line diets, the genetic modification must have had significant effects on the utilization and nutritional value of the crop. Consequently, as under these conditions the effects of the GM-material in the diet on the metabolism and health of the rats are not substantially equivalent to those of the appropriate parent line, the GM-material cannot be accepted for inclusion in the human diet without further testing. Under most circumstances this GM-material may not be acceptable in animal feedstuffs either. If the negative health effect of the GM-diet is also observed with the parent-line diet spiked with the gene product, the use of this gene in GM-food/feed is not acceptable. If on the other hand the negative effects of the GM- food are not observed with the parent line diet containing the gene product isolated from either the GM or the original plant, they are likely to be the result of the use of the particular construct or an unwanted and unforeseen effect on the genome of the gene insertion. In this case, the gene may be suitable for transfer but either another construct/promoter is (are) needed for the genetic modification with which a more suitable GM-crop plant can be constructed or the insertion effect may be minimized by purposeful selection of the best transformant out of the many obtained in the transformation. Finally, the most important principle must not be forgotten: animal testing is but a first step and its results, even when there are no indications of harm to the animals by the GM-material in the diet, will have to be eventually confirmed by human volunteer studies. As our first contact of GM-foods is in the alimentary tract, it is of paramount importance to establish whether the transferred genes or their products in the GM-food survive digestion in the gut in vivo. If so, the possibility exists that they can have damaging effects on the intestines and the gut flora, with unpredictable physiological and immunological consequences. These effects can be particularly acute in the young, elderly and sick people suffering from HIV, hepatitis or other viral diseases since the cells of these people already contain active viral material. Since there was some indication of interference with the normal development of the sexual organs in our GM-potato feeding studies carried out with male rats, it is imperative that similar studies should also be done with young female animals. This is to be followed by investigating in detail the effects on the reproductive performance of rats (or other animals) reared and maintained on GM vs non-GM diets for several generations. These are but the first steps in a strict biological testing programme which then, in view of the results obtained, can and need be further developed and improved upon. It clearly cannot be allowed that human health and the environment be jeopardised for quick financial gain by biotech companies pushing reckless technological development. The detailed protocols described in this submission underline my belief that, if there is a willingness on the part of the regulators to embrace the concept of biological testing (in addition to possible toxicological evaluation), the methods are there and available for the work to start. 4. Other issues, environment, agribusiness, consumers relevant to GM food safety. 5. Further international co-operation on scientific GM food safety issues. On these last two issues I do not regards myself an expert and therefore leave them to others more competent to deal with them. References 1. BG Hammond et al. (1996) The feeding value of soybeans fed to rats, chickens, catfish and dairy cattle is not altered by genetic incorporation of glyphosate tolerance. The Journal of Nutrition, 126, 717-727. 2. A Pusztai et al. (1999) Expression of the insecticidal bean Ó-amylase inhibitor transgene has minimal detrimental effect on the nutritional value of peas fed to rats at 30% of the diet. The Journal of Nutrition, 129, 1597-1603. 3. SWB Ewen, A. Pusztai (1999) Effects of diets containing genetically modified potatoes expressing Galanthus nivalis lectin on rat small intestine. The Lancet, 354, 1353-1354. ??