1. TWN intro
2. LACK OF SCIENTIFIC CREDIBILITY OF GM SAFETY TESTS - Ann Clark
3. GENETICALLY MODIFIED FOODS: ARE THEY A RISK TO HUMAN/ANIMAL HEALTH? - Arpad Pusztai
THIRD WORLD NETWORK INFORMATION SERVICE ON BIOSAFETY
15 September 2001
Dear friends and colleagues,
We are pleased to share with you two important articles that deal with the safety problems of genetically modified (GM) food and crops. The first article, by Ann Clark, an associate professor of plant agriculture at Guelph University, originally appeared in the Canadian newspaper, Toronto Star, on March 12, 2001. The second article, a review of all the published scientific studies relating to GM food safety by Dr. Arpad Pusztai, was published by BioScience Productions in June 2001.
Clark's article reveals the disturbing lack of scientific credibility in the biotechnology industry's claims about the safety of genetically modified organisms (GMOs). Clark points out that there is a "near absence of credible scientific evidence on which to assess environmental and food safety risks" of GM foods and crops. She cites the much publicised detailed database search by Jose Domingo which only found "a grand total of just eight refereed journal articles dealing with any aspect of the safety of GM foods. The eight included only four actual feeding trials, of which three were from Monsanto's teams."
She also highlights the fact that the report of the EU-US Biotechnology Consultative Forum states that, "There is a lack of substantial scientific data and evidence, often (presented) more as personal interpretations disguised as scientifically validated statements."
Pusztai's review shows that GM crops and food are being deemed safe for public consumption even though few scientific studies have been carried out on the risks to human and animal health. In fact, there are no peer-reviewed publications of clinical studies on the human health effects of GM food. In his review of the few tests and studies done, Pusztai concludes that current scientific data is "woefully inadequate" to assess the potential harm of GM crops and food and that the scientific quality of what has been published "is in most instances not up to expected standards".
Pusztai also reveals some frightening results from studies done on the effects of GM food on animal health. In the case of the FLAVR SAVR tomato, 7 out of 20 female rats, and none in the controls, showed stomach lesions. Despite this, no further experiments were conducted. 7 out of 40 rats fed with GM tomatoes died within two weeks. But despite these results, the FLAVR SAVR tomato was declared to be safe for human consumption. Pusztai's own research found that rats fed with GM potatoes suffered abnormal intestinal growth.
These two articles highlight the need for more and better testing of GM food and crops based on sound science. In the absence of independent and credible GM food safety tests, the precautionary principle must be applied.
With best wishes,
Website: www.twnside.org.sg TWN biosafety information service is archived on our website: www.twnside.org.sg/bio_7.htm
THE LACK OF SCIENTIFIC CREDIBILITY OF GM FOOD SAFETY TESTS
By Ann Clark
Until recently, people tended to identify most of the concern about genetically modified (GM) agriculture with groups such as the Council of Canadians, Greenpeace and Friends of the Earth. Industry proponents wasted little time in painting these people as misinformed, hysterical greenies. But thanks to those groups, informed citizen opposition has slowed adoption of GM crops to a crawl, providing much-needed breathing space for senior scientists, lawyers, and physicians to reflect upon the issues and begin to speak out.
Proponent efforts to paint the opposition as ill-informed malcontents and luddites sound increasingly silly in the face of the significant doubts now reaching the public media from prestigious scientific analysts.
One common criticism in many such studies is the near absence of credible scientific evidence upon which to assess environmental and food safety risks. Last June, the prestigious journal Science reported a detailed database search by Jose Domingo, who could find a grand total of just eight refereed journal articles dealing with any aspect of the safety of GM foods. The eight included only four actual feeding trials, of which three were from Monsanto teams.
The final report of the elite, hand-picked EU-U.S. Biotechnology Consultative Forum, which came out in December, 2000, stated, "There is a lack of substantial scientific data and evidence, often (presented) more as personal interpretations disguised as scientifically validated statements." The full report is available at http://europa.eu.int
The Royal Society of Canada just came out with a new report entitled The Future of Food Biotechnology. Elements of Precaution: Recommendations for the Future of Food Biotechnology in Canada. This group of distinguished senior scientists identified numerous critical failings in the Canadian GM regulatory process, and were particularly critical of the pivotal role accorded the unscientific concept of "substantial equivalence." The report is available at http://www.rsc.ca
In another recent issue of Science, U.S. government scientists LaReesa Wolfenbarger and Paul Phifer noted that "key experiments on both the environmental risks and benefits are lacking." They identified numerous critical deficiencies whether GM crops are indeed safe for the environment. Each of these studies calls for substantially increased research to figure out whether any risk exists, let alone how to test for such risk or to do about it.
In effect, governments have authorized the commercial release of almost 50 GM crops, which were sown over 100 million acres in 1999 (71 per cent in the U.S., 17 per cent in Argentina, and 10 per cent in Canada), and yet we still don't know enough even to identify the food safety and environmental risks, let alone tests for them.
In a nutshell, we don't know enough about basic gene function, the complexity of metabolic pathways, and the ecological implications of even modest genetic modifications to be doing what we are doing, commercially. As stated colloquially by Craig Venter, head of the Celera team that recently decoded the human genome, "We don't know s--t about biology."
With a virtual absence of refereed support for heir beliefs, industry proponents insist there is still ample evidence of the safety of GM crops, pointing to voluminous internal industry and government reports. But how credible are these reports if they are not of a sufficient caliber to be published in a refereed journal?
The requirement for publishing in a refereed journal is universally accepted in the scientific community. Authors are required to submit their work to review and critical comment from peers in the field to ensure the quality and integrity of the research. This is neither academic trivia nor overblown rhetoric, but is deadly earnest. Careers have been destroyed by this very issue, strange though it may seem.
Two years ago, Arpad Pusztai, a world-renowned authority on plant proteins and nutrition, with nearly 300 refereed publications to his credit was fired and treated disgracefully by his own colleagues for committing the unforgivable sin of speaking publicly about his concerns about GM food safety prior to publishing his findings in a refereed journal. Pusztai had conducted meticulous studies that found organ size and intestinal integrity were hurt in rats fed potatoes that had been genetically modified to include genes from snowdrop lectin. Worse yet, rats fed plain potatoes sprinkled with snowdrop lectin did not show these effects. The study suggested that the problem related to the transgenic process, not the product.
Does it seem odd to fire a scientist for expressing his concerns? Incomprehensible? Bizarre? There's more.
The same Canadian proponents who just two years ago loudly affirmed Pusztai's firing because he had not published his work in a refereed journal are now loudly proclaiming the legitimacy of unpublished internal documents promoting GM safety. You can't have it both ways. Either research must be published in refereed journals to have scientific credibility, as was Pusztai's eventually, or not.
And if not, if unpublished internal reports are to be accepted as credible and authoritative scientific information, one must conclude that the shameless destruction of Pusztai's career and he termination of his entire research program had little to do with refereed journal publishing, and everything to do with what he found.
GENETICALLY MODIFIED FOODS: ARE THEY A RISK TO HUMAN/ANIMAL HEALTH?
By Arpad Pusztai, Ph.D.
Scarcity of safety tests
How can the public make informed decisions about GM foods when there is so little information about its safety? The lack of data is due to a number of reasons, including:
* It's more difficult to evaluate the safety of crop-derived foods than individual chemical, drug, or food additives. Crop foods are more complex and their composition varies according to differences in growth and agronomic conditions.
* Publications on GM food toxicity are scarce. An article in Science magazine said it all: "Health Risks of Genetically Modified Foods: Many Opinions but Few Data".1 In fact, no peer-reviewed publications of clinical studies on the human health effects of GM food exist. Even animal studies are few and far between.
* The preferred approach of the industry has been to use compositional comparisons between GM and non-GM crops. When they are not significantly different the two are regarded as "substantially equivalent", and therefore the GM food crop is regarded as safe as its conventional counterpart. This ensures that GM crops can be patented without animal testing. However, substantial equivalence is an unscientific concept that has never been properly defined and there are no legally binding rules on how to establish it.2
When food-crops are genetically modified, ("genetically modified" food is a misnomer!) one or more genes are incorporated into the crop's genome using a vector containing several other genes, including as a minimum, viral promoters, transcription terminators, antibiotic resistance marker genes and reporter genes. Data on the safety of these are scarce even though they can affect the safety of the GM crop. For example:
* DNA does not always fully break down in the alimentary tract.3,4 Gut bacteria can take up genes and GM plasmids5 and this opens up the possibility of the spread of antibiotic resistance.
* Insertion of genes into the genome can also result in unintended effects, which need to be reduced/eliminated by selection, since some of the ways the inserted genes express themselves in the host or the way they affect the functioning of the crop's own genes are unpredictable. This may lead to the development of unknown toxic/allergenic components, which we cannot analyze for and seriously limiting the selection criteria.
Currently, toxicity in food is tested by chemical analysis of macro/micro nutrients and known toxins. To rely solely on this method is at best inadequate and, at worst, dangerous. Better diagnostic methods are needed, such as mRNA fingerprinting, proteomics and secondary metabolite profiling.6 However, consuming even minor constituents with high biological activity may have major effects on the gut and body's metabolism, which can only be revealed from animal studies. Thus novel toxicological/nutritional methods are urgently needed to screen for harmful consequences on human/animal health and to pinpoint these before allowing a GM crop into the food chain.7
Safety tests on commercial GM crops
GM tomatoes: The first and only safety evaluation of a GM crop, the FLAVR SAVR[TM] tomato, was commissioned by Calgene, as required by the FDA. This GM tomato was produced by inserting kanr genes into a tomato by an 'antisense' GM method. The test has not been peer-reviewed or published but is on the internet.8 The results claim there were no significant alterations in total protein, vitamins and mineral contents and in toxic glycoalkaloids.9 Therefore, the GM and parent tomatoes were deemed to be "substantially equivalent."
In acute toxicity studies with male/female rats, which were tube-fed homogenized GM tomatoes, toxic effects were claimed to be absent. In addition, it was concluded that mean body and organ weights, weight gains, food consumption and clinical chemistry or blood parameters were not significantly different between GM-fed and control groups. However:
* The unacceptably wide range of rat starting weights (±18% to ±23%) invalidated these findings.
* No histology on the intestines was done even though stomach sections showed mild/moderate erosive/necrotic lesions in up to seven out of twenty female rats but none in the controls. However, these were considered to be of no importance, although in humans they could lead to life-endangering hemorrhage, particularly in the elderly who use aspirin to prevent thrombosis.
* Seven out of forty rats on GM tomatoes died within two weeks for unstated reasons.
* These studies were poorly designed and therefore the conclusion that FLAVR SAVR[TM] tomatoes were safe does not rest on good science, questioning the validity of the FDA's decision that no toxicological testing of other GM foods will in future be required.
GM maize: Two lines of Chardon LL herbicide-resistant GM maize expressing the gene of Phosphinothricin Acetyltransferase Enzyme (PAT-PROTEIN) before and after ensiling showed significant differences in fat and carbohydrate contents compared with non-GM maize and were therefore substantially different. Toxicity tests were only performed with the PAT-PROTEIN even though with this the unpredictable effects of the gene transfer or the vector or gene insertion could not be demonstrated or excluded. The design of these experiments was also flawed because:
* The starting weight of the rats varied by more than ± 20% and individual feed intakes were not monitored.
* Feed conversion efficiency on PAT-PROTEIN was significantly reduced.
* Urine output increased and several clinical parameters were also different.
* The weight and histology of the digestive tract (and pancreas) was not measured.
Thus, GM maize expressing PAT-PROTEIN may present unacceptable health risks.
GM soybeans: To make soybeans herbicide resistant, the gene of 5-enolpyruvylshikimate-3-phosphate synthase from Agrobacterium was used.
Safety tests claim the GM variety to be "substantially equivalent" to conventional soybeans.10 The same was claimed for GTS (glyphosate-resistant soybeans) sprayed with this herbicide.11 However, several significant differences between the GM and control lines were recorded10 and the statistical method used was flawed because:
* Instead of comparing the amounts of components in a large number of samples of each individual GTS with its appropriate parent line grown side-by-side and harvested at the same time, the authors compared samples from different locations and harvest times.
* There were also differences in the contents of natural isoflavones (genistein, etc.) with potential importance for health.12
* Additionally, the trypsin inhibitor (a major allergen) content was significantly increased in GTS.10
Because of this, and the large variability (± 10% or more), the lines could not be regarded as "substantially equivalent."
GM potatoes: There is only one peer-reviewed publication on GM potatoes that express the soybean glycinin gene.13 However, the expression level was very low and no improvements in the protein content or amino acid profile were obtained.
GM rice: The kind that expresses soybean glycinin gene (40-50 mg glycinin/g protein) has been developed14 and is claimed to contain 20% more protein. However, the increased protein content was probably due to a decrease in moisture rather than true increase in protein putting a question mark over the significance of this GM crop.
GM cotton: Several lines of GM cotton plants have been developed using a gene from Bacillus thuringiensis subsp. kurstaki providing increased protection against major lepidopteran pests. The lines were claimed to be "substantially equivalent" to parent lines15 in levels of macronutrients and gossypol, cyclopropenoid fatty acids and aflatoxin levels were less than those in conventional seeds. However, because of the use of inappropriate statistics it is questionable whether the GM and! non-GM lines were truly equivalent, particularly as environmental stresses could have unpredictable effects on antinutrient/toxin levels.16
Herbicide-resistant soybean: Studies have been conducted on the feeding value17 and possible toxicity18 for rats, broiler chickens, catfish and dairy cows of two GM lines of glyphosate-resistant soybean (GTS). The growth, feed conversion efficiency, catfish fillet composition, broiler breast muscle and fat pad weights and milk production, rumen fermentation and digestibilities in cows were claimed to be similar for GTS and non-GTS. However:
* These experiments were poorly designed since the high dietary protein concentration and the low inclusion level of GTS could have masked any GM effect.
* No individual feed intakes, body or organ weights were given and no histology was performed, except some qualitative microscopy on the pancreas.
* The feeding value of the two GTS lines was not substantially equivalent either because the rats grew significantly better on one of the GTS lines than on the other.
* The experiment with broiler chicken was a commercial and not a scientific study.
* The catfish experiment showed again that the feeding value of one of the GTS lines was superior to the other.
* Milk production and performance of lactating cows also showed significant differences between cows fed GM and non-GM feeds.
* Moreover, testing of the safety of 5-enolpyruvylshikimate-3-phosphate synthase which renders soybeans glyphosate-resistant18 was irrelevant because in the gavage studies an E. coli recombinant and not the GTS product was used. Their effects could be different as the differences in post-translational modification could have impaired their stability to gut proteolysis.
Thus, the claim that the feeding value of GTS and non-GTS lines was substantially equivalent is at best premature.
In a separate study19 it was claimed that rats and mice which were fed 30% toasted GTS or non-GTS in their diet had no significant differences in nutritional performance, organ weights, histopathology and production of IgE and IgG antibodies. However, under the unphysiological -- basically, starvation -- conditions of these experiments when, instead of the normal daily growth of 5-8 g per day, the rats grew less than 0.3 g and mice not at all, no valid conclusions could be drawn.
GM corn: One broiler chicken feeding study with rations containing transgenic Event 176 derived Bt corn (Novartis) has been published.20 However, the results of this trial are more relevant to commercial than academic scientific studies.
GM peas: The nutritional value of diets containing GM peas expressing bean alpha-amylase inhibitor when fed to rats for 10 days at two different (30% or 65%) dietary inclusions, was shown to be similar to that of parent-line peas.21
* Even at 65% level the difference was small mainly because the alpha-amylase inhibitor expressed in the peas was quickly digested in the rat gut and its antinutritive effect abolished. Unfortunately no gut histology was done or lymphocyte responsiveness measured.
* Although some organ weights, mainly the caecum and pancreas were different, those of others were remarkably similar suggesting that GM peas may be used in the diets of farm animals at low/moderate levels if their progress was carefully monitored.
However, to establish its safety for humans a more rigorous specific risk assessment will have to be carried out with several GM lines. This should include:
* An initial nutritional/toxicological testing on laboratory animals
* If no harmful effects are then detected, it should be followed by clinical, double-blind, placebo-type tests with human volunteers, keeping in mind that any possible harmful effects would be particularly serious with the young, old, and disabled.
A protocol for such testing was given at the OECD conference in Edinburgh, February 2000 and subsequently published.22
GM potatoes: In a short feeding study to establish the safety of GM potatoes expressing the soybean glycinin gene, rats were daily force-fed with 2 g of GM or control potatoes/kg body weight.23 Although no differences in growth, feed intake, blood cell count and composition and organ weights between the groups was found, the potato intake of the animals was too low and unclear, whether the potatoes were raw or boiled.
Feeding mice with potatoes transformed with a Bacillus thuringiensis var. kurstaki Cry1 toxin gene or the toxin itself was shown24 to have caused villus epithelial cell hypertrophy and multinucleation, disrupted microvilli, mitochondrial degeneration, increased numbers of lysosomes and autophagic vacuoles and activation of crypt Paneth cells. The results showed that despite claims to the contrary, CryI! toxin was stable in the mouse gut and therefore GM crops expressing it need to be subjected to "thorough tests...to avoid the risks before marketing.24
In another study, young, growing rats were pair-fed on iso-proteinic and iso-caloric balanced diets containing raw or boiled non-GM potatoes and GM potatoes with the snowdrop (Galanthus nivalis) bulb lectin (GNA) gene.25 The results showed that the mucosal thickness of the stomach and the crypt length of the intestines of rats fed GM potatoes was significantly increased. Most of these effects were due to the insertion of the construct and not to GNA which had been been pre-selected as a non-mitotic lectin unable to induce hyperplastic intestinal growth26 and epithelial T lymphocyte infiltration. Although there is controversy about the tests, most of the adverse comments on this Lancet paper were personal, non-peer reviewed opinions and, as such, of limited scientific value. The findings, on the other hand, were published in a peer-reviewed publication25 and the criticism replied to.7 The work, however, has not been repeated nor results contradicted and it is therefore imperative that the effects on the gut structure and metabolism of all other GM crops developed using similar techniques and genetic vectors should be thoroughly investigated before their release into the food chain.
GM tomatoes: This study with a GM tomato expressing B. thuringiensis toxin CRYIA(b) gene was published in a book and not in a peer-reviewed journal. However, its importance was underlined by the immunocytochemical demonstration of in vitro binding of Bt toxin to the caecum/colon from humans and rhesus monkeys.27 Although in vivo the Bt toxin was not bound by the rat gut, this was possibly due to the authors' use of recombinant Bt toxin.
One of the major health concerns with GM food is its potential to increase allergies and anaphylaxis in humans eating unlabeled GM foodstuffs.
* When the gene is from a crop of known allergenicity, it is easy to establish whether the GM food is allergenic using in vitro tests, such as RAST or immunoblotting, with sera from individuals sensitised to the original crop. This was demonstrated in GM soybeans expressing the brasil nut 2 S protein28 or in GM potatoes expressing cod protein genes.29
* It is also relatively easy to assess whether genetic engineering affected the potency of endogenous allergens.30 Some farm workers exposed to B. thuringiensis pesticide were shown to have developed skin sensitization and IgE antibodies to the Bt spore extract. With their sera it may now therefore be possible to test for the allergenic potential of GM crops expressing Bt toxin.31 It is all the more important because Bt toxin Cry1Ac has recently been shown to be a potent oral/nasal antigen and adjuvant.32
Assessment of the allergenicity of a GM foodcrop, however, is difficult when the gene is transferred from a source not eaten before or with unknown allergenicity or on gene transfer/insertion a new allergen or adjuvant is developed or the expression of a minor allergen is increased. Unfortunately, while there are good animal models for nutritional/toxicological testing, no such models exist for allergenicity testing.
* Presently only indirect and rather scientifically unsound methods, such as finding SHORT sequence homologies (at least 8 contiguous amino acids) to any of the about 200 known allergens, are used for the assessment of allergenicity.
* The decision-tree type of indirect approach based on factors (such as size and stability) of the transgenically expressed protein33 is even more unsound, particularly as its stability to gut proteolysis is assessed by an in vitro (simulated) testing34 instead of in vivo (human/animal) testing and this is fundamentally wrong. The concept that most allergens are abundant proteins is also misleading because for example Gad c 1, the major allergen in codfish, is not a predominant protein.29
* However, when the gene responsible for the allergenicity is known, such as the gene of the alpha-amylase/trypsin inhibitors/allergens in rice, cloning and sequencing opens the way for reducing their level by antisense RNA strategy.35
Thus, in the absence of reliable methods for allergenicity testing, it is at present impossible to definitely establish whether a new GM crop is allergenic or not before its release into the human/animal food/feed chain.
One has to agree with the piece in Science1 that there are many opinions but scarce data on the potential health risks of GM food crops, even though these should have been tested for and eliminated before their introduction. Our present data base is woefully inadequate.
Moreover, the scientific quality of what has been published is, in most instances not up to expected standards. If, as claimed, our future is dependent on the success of the promise of genetic modification delivering wholesome, plentiful, more nutritious and safe GM foods, the inescapable conclusion of this review is that the present crude method of genetic modification has so far not delivered these benefits and the promise of a superior second generation is still in the future.
Although it is argued by some that small differences between GM and non-GM ! crops have little biological meaning, it is clear that most GM and parental line crops fall short of the definition of "substantial equivalence." In any case, this crude, poorly defined and unscientific concept outlived its possible previous usefulness and we need novel methods and concepts to probe into the compositional, nutritional/toxicological and metabolic differences between GM and conventional crops and into the safety of the genetic techniques used in developing GM crops if we want to put this technology on a proper scientific foundation and allay the fears of the general public. We need more science, not less.6,7
(c) 2001, BioScience Productions, Inc., an organization promoting bioscience literacy. Educators have permission to reprint articles for classroom use; other users, please contact editor for reprint permission.
The full references will be placed on our website: http://www.twnside.org.sg/bio_7.htm