Scientific Queries on Health Hazards of GE Foods

Some of the most crucial scientific questions concerning health effects of GE and GEOs (genetically engineered organisms) were raised up to twenty years ago. Most of them have still not been answered at all, or have found unsatisfactory answers. We believe, as Mayer and Stirling said, "in the end it is often the case that those who choose the questions determine the answers". Will another twenty years pass before societies realize the urgent need for public funding of genuinely independent risk- and hazard-related research? The time for such investment is now so that a new scientific culture with working hypotheses rooted in the Precautionary principle (PP) can discover other, possibly even more important questions of safety.

In the present article we will mainly confine ourselves to putative health hazards related to GE plants (GEPs) used as food or feed, with some brief notes on GE vaccines as well as the novel si RNA- and nanobio-technologies. This does not mean that we do not recognize the paramount, indirect threats to public health posed by social, cultural, ethical, economic and legal issues.

In the specific context of food or feed safety assessment "hazard" may be defined as a biological, chemical or physical agent in, or condition of, food with the potential to cause an adverse health effect. The hypothetical hazards of whole GM foods, i.e. those hazards that have been realized so far, fall into a few broad categories. They are either related to the random and inaccurate integration of transgenes into recipient plant genomes, uncertainty with regard to direct or indirect effects of the polypeptide product of the transgene, or uncertainty with regard to DNA types and circumstances promoting uptake and organ establishment of foreign DNA from mammalian gastro-intestinal tracts .

A number of scientific concerns have been raised in connection with public and animal health. In the following we will discuss, in some detail, a few of these. Some of them have been thoroughly discussed in excellent, very recent reviews .

Our contribution is based on "gene ecology"; a new, cross-disciplinary scientific field intended to provide holistic knowledge based on the precautionary principle .

Some of the concerns we raise will also be relevant for environmental risk assessments of GEOs due to the fact that the processes discussed can take place in an ecosystem at large as well as in the ecosystems represented by mammalian organisms.

Do we know that any GE food/feed is safe for consumption?

For a composite material like food/feed, reductionistic approaches testing single components in vitro are highly unsatisfactory and cannot by definition clarify important safety issues. In spite of the obvious need, very few studies designed to investigate putative effects of GE nucleic acids or food/feed on potential animal or human consumers have been published in peer-reviewed journals. A consensus has emerged that the effects observed in some published studies must be experimentally followed up. To this day, this has not been done.

Most of the animal feeding studies performed so far have been designed exclusively to reveal husbandry production differences between GEOs and their unmodified counterparts. Studies designed to reveal physiological or pathological effects are extremely few, and they demonstrate a quite worrisome trend : Studies performed by the industry find no problems, while studies from independent research groups often reveal effects that should have merited immediate follow-up, confirmation and extension.

Such follow-up studies have not been performed. There are two main factors accounting for this situation: The lack of funds for independent research, and the reluctance of producers to deliver GE materials for analysis.

Can we rely on the transgenic DNA sequences given by GE food/feed producers?

If the transgenic DNA sequences given in the notifications differ from the inserted sequences found in the GEPs, the risk assessments made prior to approval of the GEPs for marketing do not necessarily cover the potential risks associated with the GEPs.

The most thoroughly studied transgenic events are:

Bt-transgenic maize Mon810 Bt- and glufosinate-transgenic maize Bt176 Glyphosate-transgenic maize GA21 Glufosinate-transgenic maize T25 (Liberty Link) Glyphosate-transgenic soybean GTS 40-3-2

Even amongst the most thoroughly studied and some of the oldest commercial GEPs, recent independent work has revealed that the nature of the rearrangements vary, and deletions (Mon810, GA21, Bt176), recombinations (T25, GTS 40-3-2, Bt176), tandem or inverted repeats (T25, GA21, Bt176) as well as rearranged transgenic fragments scattered through the genome (Mon810) have been reported.

The transgenic modification techniques are prone to introduce such rearrangements because exogenous DNA transfer in plants elicits a "wound" response, which activates nucleases and DNA repair enzymes. This may result in either degradation of the incoming DNA, or insertion of rearranged copies into the plant DNA . In addition, the nature of the DNA constructs used to make transgenic plants may influence the rearrangement tendencies for a given transgenic event. Some genetic elements in the constructs may act as "hotspots" and elicit recombinations at high frequencies.

While it was earlier assumed that integration of transgenic constructs took place at random locations in the recipient plant genome, it has now become apparent that integration sites are concentrated in or near elements such as retrotransposons (T25, Mon810, GA21) and repeated sequences (Bt11 maize) , and this poses additional risks. Firstly, by introducing a new promoter or new enhancer motifs, transgenic insertions into, or close to, such elements may lead to altered spatial and temporal expression patterns of plant genes located close to and even far from, the insert. Secondly, a strong retrotransposon LTR promoter may upregulate the transgene expression level. Thirdly, defective retrotransposons may start "jumping" under the influence of transacting factors recruited by the insert . All these events may have unpredictable effects on the long-term genetic stability of the GEOs, as well as on their nutritional value, allergenicity and toxicant contents. These putative processes represent areas of omitted research with regard to health effects of GEOs.

Are transgenic DNA and proteins taken up from mammalian GIT (gastro-intestinal tracts)?

If DNA and proteins from GEOs persist in, and are taken up from mammalian GIT, this could theoretically, as will be further explained below, ultimately lead to development of chronic disease conditions. The fate and consequences of DNA persistence and uptake is, however, not extensively studied, and therefore represents yet another area of uncertainties connected to GEPs.

It has generally been claimed that DNA and proteins are effectively degraded in mammalian GITs. This has been based on assumptions that have never been systematically examined. A restricted number of recent publications have demonstrated that foreign DNA and also proteins may escape degradation, to persist in the GIT and even to be taken up from the intestines and transported by the blood to internal organs in biologically meaningful versions. These findings should not have come as such a surprise, since scientific articles from the 1990s strongly indicated that this was an area of omitted research, as stated by a number of reports .

Briefly summarised, the present conception of DNA persistence and uptake includes long fragments of ingested DNA. DNA may be detected in the feces, the intestinal wall, peripheral white blood cells, liver, spleen and kidney, and the foreign DNA may be found integrated in the recipient genome. When pregnant animals are fed foreign DNA, fragments may be traced to small cell clusters in fetuses and newborns. The state of GIT filling, and the feed composition may influence DNA persistence and uptake. Complexing of DNA with proteins or other macromolecules may protect against degradation.

So far only two published reports have investigated the fate of foreign/transgenic DNA in humans. The consequences of DNA persistence and uptake thus represent yet another area of omitted research. Extrapolating from a number of experiments in mammalian cell cultures and in experimental animals, it is conceivable that in some instances insertion of foreign DNA may lead to alterations in the methylation and transcription patterns of the recipient cell genome, resulting in unpredictable levels of gene expression levels and products. Furthermore, even small inserts may result in a so-called "destabilisation" process, the end-point of which may be malignant cancer cells .

The BSE/new variant Creutzfeld-Jacob's Disease epidemics caused by the prion proteins painfully illustrated the phenomenon of protein persistence, uptake and biological effects.

Two recent publications indicate that this phenomenon may be more general that realized. A hallmark of prion diseases and a number of other debilitating, degenerative diseases, i.e. Alzheimer's and Huntington's diseases, is deposition of "amyloid fibrils". Recent studies indicate that any protein can adopt a confirmation known as "amyloid" upon exposure to appropriate environmental conditions. Whether that is the case for GE food/feed that is already in the marketplace is unknown.

The consequences of protein persistence and uptake will vary with the given situation. Generally spoken there is a possibility that toxic, immunogenic/allergenic or carcinogenic molecules may gain entry to the organism via cells in the gastrointestinal walls. The persistence of the Bt-toxin Cry1Ab in feces means a potential for spread on the fields through manure. The ecological effects, e.g. on insect larvae and earthworms, are at the moment an issue of shear speculation.

Have the protein contents of GE food been altered in unpredictable ways?

Transgenes or upregulated plant genes may give rise to toxicants, anti-nutritients, allergens and, putatively, also carcinogenic or co-carcinogenic substances. The concentration of a given transgenic protein may vary according to the location(s) in the recipient host cell genome of inserted GE construct DNA, and to environmental factors influencing the activity of the transgenic regulatory elements, e.g. the 35S CaMV promoter. The biological effects of a given transgenic protein, e.g. the Cry1Ab Bt-toxin, may be unpredictably influenced by posttranslational modifications, alternative splicings, alternative start codons for transcription, chimeric reading frames resulting from integration into the reading frame of a plant gene, and complex formation with endogenous plant proteins.

The influence of foreign DNA insertion on endogenous plant gene expression patterns may vary with local environmental factors, the actual insertion site(s), the number and stability of the inserts, transgenic promoter effects, methylation patterns of the insert(s), and post-transformational mutations in the transgenic protein coding as well as in regulatory sequences. Even a single nucleotide change may affect the properties of a protein, or it may create a new transcription factor-binding motif. Detailed studies of these phenomena under authentic conditions are lacking, and hence we are confronted with yet another area of omitted research.

May GE food/feed give allergies?

One of the major health concerns related to GEPs is that the transgenic product itself, e.g. a Bt toxin, or changed expression of endogenous plant genes may result in allergenic compounds. The risk assessment of allergens often follows an allergenicity decision tree . These "trees" are based on in vitro tests comparing a limited number of structures, usually only one, of the transgenic protein with known allergens. Hence, these comparisons are hopeful that the protein isolated for the test matches all proteins produced from the same gene in the GEP. But in fact this is unlikely because allergenicity tests are usually carried out with bacteria-, not in planta-produced versions of the transgenic protein. Glycosylation invariably takes place in plants, but not in bacteria, so this form of post-translational modification of both the transgenic protein and endogenous proteins would not be tested. Allergenic characteristics of proteins, and also their resistance to degradation in the organism, can be affected by glycosylation. Other protein modifications may also take place, adding to the unpredictability of transgenic products .

Another important question related to allergenicity is whether post marketing surveillance can provide useful information about allergens in GE foods. For a number of reasons this is not likely to happen . Treatment of allergy is symptomatic, whatever the cause may be. The allergic case is often isolated, and the potential allergen is rarely identified. The number of allergy-related medical visits is not tabulated. Even repeated visits due to well-known allergens are not counted as part of any established surveillance system. Thus, during the October 2000 Starlink episode, it proved very difficult to evaluate Starlink (containing Bt-toxin Cry9C) as a human allergen . An additional reason for this was that the ELISA tests, used by FDA, that found no anti-Cry9C antibodies in suspected human cases were dubious because bacterial, recombinant antigens were used instead of the Cry9C maize versions that the individuals had been exposed to.

Case: Bt toxins in Bt-transgenic GEPs

It is very important to be aware of the fact that the Bt-toxins expressed in GEPs have never been carefully analysed, and accordingly, their characteristics and properties are not known. What is clear from the starting point, however, is that they are vastly different from the bacterial Bacillus thuringiensis protoxins, used in organic and traditional farming and forestry for decennia . The difference is evident already at the gene level, since the versions found in GEOs are engineered to produce active Bt toxins. By extrapolation these have a number of potentially unwanted biological characteristics, ranging from solubilization of the protein under natural conditions and effects on insect and mammalian cells, to persistence and non-target effects in the environment. In addition, the posttranslational modifications that may influence conformations, cellular targets and biological effects of GEP-expressed Bt-toxins are unknown, and hence we once more identify an area of omitted research.

During the last few years a number of observations that may be conceived of as "early warnings" of potential health and environmental risks, have appeared in the literature. Most of them have, however, not been followed up by extended studies.

Case: Transgenic, glyphosate-tolerant (Roundup Ready) GEPs

These GEPs have an inserted transgene, cp4 epsps, coding for an enzyme that degrades the herbicide glyphosate. The whole idea is of course the combined use of the GEP and the herbicide. Recent studies indicate that in some cases such GEPs are associated with greater usage of glyphosate than the conventional counterparts . A very restricted number of experimental studies have been devoted to health or environmental effects of the GEPs or the herbicide itself. Some of these may be considered "early warnings" of potential health and environmental risks, and they should be rapidly followed up to confirm and extend the findings . Consequently: yet another area of omitted research.

Is the 35S CaMV promoter inactive in mammalian cells?

Cauliflower mosaic virus (CaMV) is a DNA-containing para-retrovirus replicating by means of reverse transcription (Poogin et al., 2001). One of the viral promoters, called 35S is a general, strong plant promoter. It has been used to secure expression of the transgenes in most of the GEOs commercialized so far.

Industry proponents have claimed unconditionally that the 35S is an exclusive plant promoter, and hence cannot, even theoretically, represent a food/feed safety issue .

In addition to studies in yeast and in Schizosaccharomyces pombe , there are published studies indicating that the 35S CaMV promoter might have potential for transcriptional activation in mammalian systems . And the final proofs have been made available during the last couple of years. First, 35S promoter activity was demonstrated in human fibroblast cell cultures , thereafter in hamster cells , and very recently one of us (TT) has demonstrated substantial 35S promoter activity in human enterocyte-like cell cultures. Such cells are lining up the surface of human intestines. However, no published studies have investigated 35S CaMV activity in vivo, and this is hence an obvious area of omitted research.

May the use of antibiotic resistance marker genes (e.g. nptII) present health hazards?

The antibiotic kanamycin is used extensively in crop genetic engineering as a selectable marker, inter alia in GE oilseed rape event lines like MS1Bn x RF1Bn and Topas 19/2.

A selectable marker is a gene inserted into a cell or organism to allow the modified form to be selectively amplified while unmodified organisms are eliminated. In crop genetic engineering the selectable marker is used in the laboratory to identify cells or embryos that carry the genetic modifications that the engineer wishes to commercialize. The selection gene is used once briefly in the laboratory, but thereafter the genetically modified (GM) crop has the unused marker gene in each and every one of its cells.

There are multiple well-known mechanisms for cross-resistance to antibiotics of a particular type . Kanamycin is a member of the family aminoglycoside antibiotics. There are approximately 17 different classes of aminoglycoside-modifying enzymes. Some of these inactivate up to four different aminoglycosides. Cross-resistance between kanamycin and other aminoglycosides, e.g. gentamycin and tobramycin, was found to vary markedly between isolates. All of the antibiotics mentioned are used to treat human diseases.

Along with cross-resistance to aminoglycoside antibiotics, pathogenic bacteria frequently develop multiple drug resistance transmitted on a single plasmid . Pathogenic bacteria do acquire plasmids with multiple antibiotic resistance genes in areas where the antibiotics are used extensively. Such incidents illustrate the potential health effects of HGT. Multiple resistance genes on a single plasmid can simultaneously adapt a bacterium to several unrelated antibiotics. One antibiotic at a time is all that is necessary to maintain the plasmid.

In spite of the belief of many genetic engineers that kanamycin is no longer employed in medical applications, there is evidence that the antibiotic is used extensively for some applications.

Concluding remarks: Where do we go from here?

We have discussed in some detail a handful of selected, unanswered risk questions related to the first generation of transgenic GEOs. There are many more risk issues. Among them are issues of Horizontal Gene Transfer (HGT), the new generations of multitransgenic GEOs for pharmaceutical and industrial purposes, safety questions related to GE vaccines, the new nanobiotechnology approaches and the applications of small inhibitory (si) RNAs for a number of medical purposes. Furthermore, we have the "questions not yet asked", and we have the problem of whether available methods and regulatory frameworks will be able to pick up and manage the conceived risks once they become reality.

In recent publications it has been demonstrated that the presently used sampling and detection methods may fail to detect GE materials in food and feed. In another article it was demonstrated that HGT events, that potentially carry very serious public health consequences, would not be detected in time for any meaningful preventive actions. And it has been illustrated that the siRNA techniques are not as "surgically targeted" as initially indicated.

We are left with a high number of risk issues lacking answers, adding up to a vast area of omitted research, and this falls together in time with a strong tendency towards corporate take-over of publicly funded research institutions and scientists.

We must as citizens and professionals join together to reverse the present situation. Publicly funded, independent research grants must become a hot political issue. That would be the most efficient remedy for lacking answers and corporate take-over of science. And finishing off, we once more quote Mayer and Stirling : "Deciding on the questions to be asked and the comparisons to be made has to be an inclusive process and not the provenance of experts alone". But then again, whom should the society rely on for answers and advice when the time comes that all science resource persons work directly or indirectly for the GE producers?

Dr. Terje Traavik is the author of more than 180 scientific articles and book chapters. He founded and was the professor of virology at the University of Troms�, Norway from 1983-2003. He has had a high number of national and international assignments. At the present he serves as the Executive Committee Chairman for the GE/GMO Biosafety Capacity Building Program covered by a MoU (Memorandum of Understanding) between GENOK-Norwegian Institute of Gene Ecology and and UNEP (United Nations Environment Program). Being originally a medical and molecular virologist, Traavik later crossed into molecular and cellular cancer research. In 1992 he received the Erna and Olav Aakre Foundation Prize for Excellent Cancer Research. In the early 1990s he was the Board Chairman of the national research program "Environmental effects of biotechnology", which was funded by the Research Council of Norway. In 1997 he initiated and became the first Director of GENOK-Norwegian Institute of Gene Ecology, and since 2003 he is professor of gene ecology at the University of Tromso.

Dr. Jack Heinemann is at the present an Assoc. Professor at the School of Biological Sciences, University of Canterbury, Christchurch. He is the Director of the NZIGE-New Zealand Institute of Gene Ecology, and an adjunct professor at GENOK-Norwegian Institute of Gene Ecology. He serves on the United Nations Environment Programme-GENOK Biosafety Capacity Building Executive Committee. Dr. Heinemann was the 2002 recipient of the New Zealand Association of Scientists Research Medal. He is the author of a high number of scientific articles, reviews and book chapters. He was one of the real pioneers within HGT (horizontal gene transfer) research, and has given major contributions to this area as well as related fields within bacterial genetics and molecular biology.