By Dr Michael Antoniou, Senior Lecturer in Molecular Genetics, GKT Medical School, Guy's Hospital, London, UK

The greatest claim of those who endorse the use of genetic engineering or modification (GM) in agriculture, is that it is not only a natural extension of traditional breeding methods for the production of new varieties of crops and farm animals, but it is also more precise and safer. It is said that GM:

(i) simply gives Nature a "nudge" speeding it along a pathway that it would take anyway; (ii) by moving a single gene between organisms, the outcomes of GM are more predictable (and therefore more precise and safer) than what occurs in traditional methods;

(iii) in molecular chemical terms, DNA is the same in all organisms and therefore poses no great danger when genes are moved between unrelated organisms (e.g. animals to plants).

However, since technically speaking traditional breeding and GM bear no resemblance to each other, how valid are these claims? Following is a discussion which tries to address this question from a fundamental genetics viewpoint.

Genetics, the study of genes, has two basic components. Firstly, there is the information content of each gene; that is, what gene carries the blueprint for which protein. Secondly, gene function or expression is extremely tightly controlled or regulated. Gene function needs to be controlled because the totality of the genetic information or DNA which is inherited, is retained in all the cells of the body. In other words, the information for the whole organism is present in every part. So, for example, the knowledge for making a kidney is present in the cells of the muscles and vice versa. This basic fact of life was most recently demonstrated with the creation of Dolly the cloned sheep. The genetic material for making Dolly was derived from an adult cell from the udder of a ewe clearly showing that the genetic information for making a whole sheep was present in this udder cell.

The correct genes being switched on at the right time and within appropriate cells, ensures that the correct protein and therefore appropriate structure and function, is present in the right place, time and quantity in the body. In order to achieve this, life has evolved sets of sophisticated on-off switches to drive expression of genes. It would not only be wasteful but potentially disastrous for genes carrying the information for proteins needed specifically in the liver to be switched on in the cells of the brain.

Genetic switches are now known to be organised in a hierarchical fashion. First and most important, genes are organised in distinct groups or families within the DNA in structures called "chromatin domains" (see Felsenfeld, 1992; Dillon and Grosveld, 1994; Pikaard, 1998; Paul and Ferl, 1998; Cohen et al., 2000). It is now clear that the expression and function of genes within a given chromatin domain are closely interdependent (see Wijgerde et al., 1995; Dillon et al., 1997) and that the function in one domain can influence gene functions within another distant domain (see Flavell, 1994; Matzke and Matzke, 1998). Generally, these chromatin domains can exist in two structural configurations. The DNA in a given chromatin domain can be in a tightly coiled, closed or "inactive" state in which the genes are switched off. Alternatively, a chromatin domain can be in an extended, open and "active" condition in which the genetic information is switched on and being expressed into proteins. Generally, chromatin domains are turned on and off as needed to provide an intricate, finely balanced state of gene control the complexities of which we are only just beginning to unravel. Tight gene control means, for example, that you will never find liver proteins and functions in your brain or leaf specific processes in the fruit and vice versa!

Clearly any technology, which aims to manipulate the genetic makeup of a given organism, must preserve the natural order and groupings of genes that have evolved to work harmoniously together over many millions of years. This is indeed the case with natural sexual reproduction which can only place between closely related forms of life. With sexual reproduction different variations of the same genes in their natural context (chromatin domains) are exchanged. This preserves tight control and complex interrelationships between genetic functions and their protein products that are vital for integrity of life.

In marked contrast to sexual reproduction, GM allows the isolation, cutting, joining and transfer of single or multiple genes between totally unrelated organisms circumventing natural species barriers. As a result combinations of genes are produced that would never occur naturally. GM ("transgenic") crops containing genes from viruses, bacteria, animals as well as from unrelated plants have been generated. Furthermore, the newly introduced gene units are composed of artificial combinations of genetic material. Transgenic tomatoes, potatoes and strawberries, for example, are under development which contain genes based on the "anti-freeze protein" from an arctic fish (the flounder). In addition, a promoter from a plant virus is used to allow this fish gene to switch on in its new host. All this is in turn coupled to an antibiotic resistance "marker" gene (usually from bacteria) to allow selection of the newly transformed plants. It is hoped this combination will allow greater tolerance to frost. This is clearly a great technological advance. However, the manipulation and transfer of DNA from one organism to another by GM can only be carried out with any degree of precision in lower forms of life such as bacteria and yeast although complications may arise even in these cases resulting from biochemical disturbances. The generation of transgenic plants and animals is currently an imperfect technique. Once injected into the reproductive cells of the organism, the introduced gene is randomly incorporated into the DNA of its new plant or animal host. As a result the normal genetic order is disrupted.

There is a further complication in the case of plants. The genetic engineering of plant cells is a very inefficient process. Out of the many millions of cells that are subjected to the GM transformation process, only a few will permanently assimilate the new genes. In order to help identify the transformed cells from the bulk of the non-transformed cells, the biotechnologist makes use of a gene whose presence confers resistance to a particular antibiotic. Therefore, the gene one is interested in getting into the plant cells, say for pesticide production, is linked to a separate gene for antibiotic resistance and the two are introduced together. One then grows the cells in the presence of the antibiotic. Only those cells, which have taken up the antibiotic resistance gene and, by default, the gene for pesticide production, will survive and grow. The vast majority of the cells, which do not incorporate these genes, will die. This approach is used in the production of all GM plants. As one can see this method totally depends on the function of the antibiotic resistance gene. This gene must be assimilated in a manner that will allow it to be switched on otherwise the cells will die once treated with the antibiotic.

As we discussed above, regions of DNA (chromatin domains) can be switched off ("inactive") or expressing genes ("active") as part of vital, normal genetic control mechanisms. Since the incorporation of the new genes into the host DNA in GM technology is a random affair totally beyond the control of the genetic engineer, the antibiotic resistance gene can be incorporated into either silent or active DNA. If the antibiotic resistance gene is incorporated into silent DNA it will not be switched on and therefore the cell will die in the presence of the antibiotic. If on the other hand the antibiotic resistance gene is assimilated in active DNA, it will be switched on and the cells will survive antibiotic treatment. However, by definition, active DNA is a region where other genes are already switched on and trying to function. The random incorporation of a foreign gene into the already active domain will therefore always risk disrupting the balanced functioning of the host genes. It was previously thought that host gene functions would only be disturbed if the foreign gene spliced into the middle of another gene or into the genetic switch region which controls it’s expression. However, it is now known that the functions of genes within a given chromatin domain are interdependent and in many cases genes within a family grouping compete for common control switches (see Dillon and Grosveld, 1994; Wijgerde et al., 1995; Cohen et al., 2000). This latest model of gene organisation and function predicts that the mere presence of another gene introduced by GM into a given chromatin domain, will compete with the host genes and disrupt their balanced function (Dillon et al., 1997; Cohen et al., 2000). Therefore, by relying on the selection of the transformed plant cells by the function of an antibiotic resistance gene, the biotechnologist in turn selects for events where the new genes have been spliced into regions of DNA where other genes are trying to function, therefore maximising the degree of disruption to normal host gene function.

In some cases the antibiotic resistance marker and the gene of interest are on separate fragments of DNA and can therefore insert into different regions of the host genome (as is the case with the Roundup Ready soya beans). However, the plant genetic control elements used to drive expression of the foreign gene (mostly the CaMV promoter), is unable to activate silent regions of DNA. Therefore, again the functioning of the transgene relies on integration into already active regions of DNA/chromatin. In addition, since selection of the transgenic plant or animal is made on the basis of looking for maximum expression of the transgene, it further selects for a site of integration within an active chromatin domain that will favour this function.

These selection procedures combine to maximise the degree of potential biochemical disturbance resulting from the disrupted gene function. Therefore, GM of animals and especially plants, always results in a loss, to a lesser or greater degree, of the tight genetic control and balanced functioning which is retained through conventional cross breeding. With GM, host genes can be silenced (rendered inactive; see Flavell, 1994; Matzke and Matzke, 1998) or inappropriately activated (Hirel et al., 1992) resulting in either a deficiency in a given protein(s) or the presence of the wrong protein(s) in the wrong place or in the wrong quantity or all these combined (e.g. see Dale et al., 1998). In addition, it is assumed that the introduced gene will behave in exactly the same way in it’s new host as it does in it’s native environment which frequently will not be the case.

These effects combine to always produce a totally unpredictable disturbance in host genetic function as well as in that of the introduced gene. These phenomena are technically called "position effects" and complicate the production of every GM crop (Dale et al., 1998) or animal (Palmiter and Brinster, 1986). Of the several tens or perhaps hundreds of individual plants or animals that will be produced with the same genes, only a few will meet the agricultural performance criteria that are being sought. This is because in each individual the foreign gene is spliced into a different location in the host DNA. Plants or animals with gross defects can always be spotted and discarded. However, subtle changes in host biochemistry that will always accompany the desired effects and which can result in the production of novel toxins, allergens as well as adversely affecting nutritional value, are on the whole ignored by the producers of these GM organisms.

The proponents of the use of GM in agriculture argue that mankind has been selecting and manipulating plant and animal food stocks for millennia and that this new technology is simply the next stage in this process. However, we have seen:

• Technically speaking, GM and sexual reproduction bear no resemblance to each other.
• GM plants and animals start out life in a laboratory culture dish.
• GM employs totally artificial units of genetic material which are introduced into plant and animal cells using chemical, mechanical or bacterial methods.
• GM always results in disruptions to the natural order of genes within the host DNA.
• GM also brings about combinations of genes that would never occur naturally.

The totally artificial nature of GM does not automatically make it dangerous. It is the imprecision in the manner by which genes are combined and the unpredictability in how the introduced gene will interact within its new environment which results in uncertainty. The balanced gene functions that have evolved together and which are preserved with traditional methods, are lost with GM. Disturbed biochemical function may in turn lead to the production of novel toxins and allergens (Mayeno and Gleich, 1994; Inose and Kousaku, 1995; Ewen and Pusztai, 1999a, b)

Genes have evolved to exist and work in families within the context of a given species. With traditional breeding, which can take place only between closely related organisms, these natural groupings of genes are preserved. Given these basic principles of life, the claim that the reductionist approach of GM, which moves one or a few genes between unrelated organisms, is a precise technology is highly questionable. What makes these assertions even more disputable is that by selecting for the function of the foreign gene and looking only at the desired agronomic performance as an end point, GM always results in a disruption in the natural genetic order of the host. Therefore, from the standpoint of the fundamental principles of genetics and the limitations in the technology, GM is neither more precise nor a natural extension of traditional cross breeding methods. If anything the opposite would appear to be true. GM violates the basic principles of genetic function while traditional methods work within and make the best use of the well established systems of genetics that have been laid down over millions of years of evolution.

Therefore GM foods possess new and unique safety considerations both in terms of health and to the environment. It would appear to be quite erroneous to view GM technology from purely an agriculture performance perspective upon which the current claims of precision and safety are based.

The documents submitted to ACRE and ACNFP by Aventis for commercial approval of Chardon LL maize, accurately describe the integrity and copy number of the CaMV-PAT gene unit that has been randomly spliced into the host plant genome. However, as is usually the case with commercial transgenic plants varieties, there is no data presented as to where within the host DNA the CaMV-PAT gene has been inserted. It is therefore not possible to know which host gene function(s) have possibly been disturbed and what biochemical processes potentially disrupted.

The biochemical compositional data of Chardon LL as described again in the documents submitted by Aventis (and referred to by others in this hearing), clearly show that this GM variety of maize is substantially non-equivalent to the parental strain from which it was derived. Chardon LL shows statistically significant different levels of sugars, amino acids and lipids when compared to the parental line. A plausible explanation for these unpredictable and unintended outcomes is that they are caused by the random insertion of the CaMV-PAT gene within an active region of the genome resulting in, as described above, the disruption of the balanced functioning of one or more host genes. This in turn could result in the observed biochemical disturbances. The degree of the biochemical perturbations that have been observed to take place raises major concerns that novel toxins and allergens may have also been produced which, as already highlighted by Drs Howard and Orskov during the presentation by Friends of the Earth, given the paucity and inadequacy of the tests conducted thus far it is impossible to conclude as to whether this is case or not.

The consumption of Chardon LL therefore poses major unknown health risks to both animals and humans.

References

Cohen, BA, Mitra, RD, Hughes, JD and Church, GM (2000) A computational analysis of whole-genome expression data reveals chromosomal domains of gene expression. Nat. Genet. 26: 183-186.

Dale P.J., Al-Kaff N., Bavage A., Irwin J., Senior I. (1998) Transgene expression and stability in Brassica. ACTA Horticulturae 459: 167-171.  

Dillon, N. and Grosveld, F. (1994) Chromatin domains as potential units of eukaryotic gene function. Curr. Opin. Genet. Develop. 4: 260-264.

Dillon, N. et al. (1997) The effect of distance on long-range chromatin interactions. Molecular Cell 1: 131-139.

Ewen SW and Pusztai A (1999a) Effect of diets containing genetically modified potatoes expressing Galanthus nivalis lectin on rat small intestine. Lancet 354: 1353-1354.

Ewen SW, Pusztai A (1999b) Health risks of genetically modified foods. Lancet 354: 684.

Felsenfeld, G. (1992) Chromatin as an essential part of the transcriptional mechanism. Nature 355: 219-224.

Flavell RB (1994) Inactivation of gene expression in plants as a consequence of specific sequence duplication. Proc. Natl. Acad. Sci USA 91: 3490-3496.

Gallie, DR (1998) Controlling gene expression in transgenics. Curr. Opin. Plant Biol. 1: 166-172.

Hirel, B., Marsolier, M.C., Hoarav, A., Hoarav, J., Brangeon, J., Shafer, R. and Verma, D.P.S. (1992) Forcing expression of a soybean root glutamine synthetase gene in tobacco leaves induces a native gene encoding cytosolic enzyme.Plant Molecular Biology 20: 207-218.

Inose T and Kousaku M (1995) Enhanced accumulation of toxic compounds in yeast cells having high glycolytic activity: a case study on the safety of genetically engineered yeast. International Journal Food Science Technology 30: 141-146.

Mandel, MA et al. (1992) Manipulation of flower structure in transgenic tobacco. Cell 71: 133-143.

Matzke, aj. and Matzke, MA (1998) Position effects and epigenetic silencing of plant transgenes. Curr. Opin. Plant Biol. 1: 142-148.

Mayeno AN and Gleich GL (1994) Eosinophilia-myalgia syndrome and tryptophan production: a cautionary tale. Trends in Biotechnology 12: 346-352.

Palmiter, R. D. and Brinster, R. (1986) Ann. Rev. Genet. 20: 465.

Paul, A.-L. and Ferl, R.J. (1998) Higher order chromatin structures in maize and arabidopsis. The Plant Cell 10: 1349-1359

Pikaard, C.S. (1998) Chromosome Topology-Organising Genes in Loops and Bounds. The Plant Cell 10: 1229-1231.

Wijgerde, M., Grosveld, F. and Fraser, P. (1995) Transcription complex stability and chromatin dynamics in vivo. Nature 377: 209-213.