Food Safety and Food Quality

Issues in Environmental Science and Technology

By R.E. Hester, R. M. Harrison

The Royal Society of Chemistry

Copyright © 2001 The Royal Society of Chemistry
All rights reserved.
ISBN: 978-0-85404-270-8

Contents

Assessing the Safety of GM Food Crops Andrew Chesson, 1,
Safety and Quality Research Priorities in the Food Industry Geoff Andrews, Alastair Penman and Chris Hart, 25,
Biosolid Recycling and Food Safety Issues Jim Wright, 43,
Mad Cows, Demented Humans and Food David M. Taylor, 79,
Natural and Synthetic Chemicals in the Diet: a Critical Analysis of Possible Cancer Hazards Lois Swirsky Gold, Thomas H. Slone and Bruce N. Ames, 95,
The MAFF Food LINK Research Programmes Christina Goodacre, 129,
Sensory Assessment of Food Qualities Peter J. Lillford, 143,
Subject Index, 163,

CHAPTER 1

Assessing the Safety of GM Food Crops

ANDREW CHESSON

1 Introduction

None of the crops currently used as food plants resemble their wild progenitors; all have been modified over the centuries to improve aspects of their quality and productivity. Selection of new traits initially was a pragmatic reaction to random mutational events that gave rise to ‘sports’, crop plants that differed in some way from the norm. With the development of an understanding of Mendelian inheritance, crosses could be made in a more deliberate and systematic manner and plant breeding and the development of new varieties became a more directed activity. The existing variation within the genetic stock proved inadequate and so irradiation and chemical methods were used to introduce random and often multiple mutations to provide greater diversity from which selections could be made and novel traits introduced. However, all such developments remained within a single species or at least a closely related group of species. Genes were modified or deleted but no new genetic material was introduced. Natural outcrossing was prevented by the highly effective physical and chemical barriers present in plants that restrict cross-fertilization and, in most instances, prevent hybridization.

It was not until the 20th century that species barriers finally were breached, allowing introduction of genetic material from a genetically unrelated plant. The crucial technological breakthrough was not the use of recombinant DNA technology, but the development of tissue culture methods. Tissue culture enabled various single-cell fusion methods to be used which overcame the natural barriers operating at the whole plant level. Such techniques are now considered to fall within the broad description of GM technology and have, retrospectively, been included in the legislation governing the use and introduction of products of GM technology. Modern recombinant DNA methods, the ability to excise a gene from a donor genome and introduce it into a recipient organism in a manner that allows its expression, has now largely made these earlier and far less selective fusion techniques redundant.

Some have argued that recombinant DNA technology applied to crops is simply a logical continuation of the manipulations applied since plants were first deliberately cultivated for food use. As such, the novel phenotypes produced are no different in principle from the varieties produced through conventional breeding and should be considered in the same manner. However, once the species barrier was crossed it became possible to introduce genetic material coding for products that have never been part and, with conventional breeding, never could be part of the food supply. For such traits, the security provided by a long history of use by humans is not available and the recognition and knowledge of safe levels of the natural toxicants present in most foods is not automatic. For this reason alone it is necessary to establish regulatory systems able to recognize any hazard introduced by foreign genes and assess the associated risks. Industry concerns that over-stringent requirements for evidence of safety will stifle development of a vital technology have limited validity. It is better to start with rigorous requirements for safety evaluation and then to relax the specifications in the light of experience, rather than having to tighten regulations once evidence of damage to human health has occurred.

The Risk Analysis Framework

Most assessors of risk base their approach on the framework set out in the 1997 FAO/WHO Consultation on Risk Management in which risk analysis is seen to consist of three interlocking components:

• Risk assessment

• Risk management

• Risk communication

The identification and characterization of possible hazards is the starting point for risk assessment which, when coupled with the likelihood of occurrence or exposure, gives a measure of risk. Whenever possible this risk is quantified, although this has proved difficult with many of the hazards suggested for GM foods because of difficulty in establishing measurable outcomes. Risk management then considers how any assessed risk could be reduced, either by removal of the hazard or by a reduction in exposure. Finally, in an ideal and wholly rational world, the process is completed by a cost–benefit analysis in which economic, social and ethical issues are considered and introductions weighed against the cost of inaction. As has become abundantly clear during the GM food debate, since this process requires value judgements, it is essential that those most affected are directly involved in this decision making process.

Recombinant DNA technology is described as a precise tool in which only the intended genes are introduced into the modified host. If this were the only issue then the process of safety assessment would be made easier since only the transgene(s) and their expressed product(s) would need to be considered. Unfortunately, while the nature of the vector and the tDNA may be closely defined, the same is not true of the manner and extent of its incorporation. With the present methods of transformation there is virtually no control over where in a genome the tDNA is introduced, how many copies are integrated, or whether introductions involve the entire gene sequence or just fragments of a gene. Since random insertions could disrupt and silence existing genes or lead to expression of otherwise non-expressed sequences, much has been made of the need to take account of possible ‘inadvertent’ and therefore unpredictable effects in any safety assessment. While this is a potential hazard, it should be recognized that unintended effects are just as much a problem in conventional breeding as with GM technology. Inadvertent effects have been documented for conventional crops on many occasions without triggering a need for a safety assessment of all new varieties being introduced.

2 GMOs already in the Food Chain

It has been estimated that some 60% of manufactured food items in American supermarkets contain genetically modified ingredients. The equivalent figure for the UK, before food manufacturers and retail outlets began deliberately to source non-GM ingredients, was somewhat lower at 40% of manufactured items. While the presence of GM ingredients in US products might have been expected, the extensive presence on the shelves of European supermarkets came as a considerable surprise to many European consumers. Few recognized the extent to which products of maize (maize starch or maize gluten) and soybean (soya protein, soya grits, soyabean oil, lecithin) were used in the food industry. Because of their importance to food and animal feed production, both were amongst the first crops to be targeted for genetic modification by the plant breeders in the USA. As can be seen in Table 1, the increase in the planting of GM maize and soybean has been rapid, with seed sales in 2000 in the USA indicating that about half of the total soybean and one-third of the total maize crop will be GM varieties. All of the GM maize and soybean met the safety standards of the American Food and Drugs Administration (FDA), who considered them as equivalent to the conventional crop and so did not require separation or identity preservation. While GM maize and soybean have not been grown to any extent in Europe, they were imported as the raw material or products mixed with conventional sources. The UK imports about 1% and Europe about 10% of the USA maize and soybean production. North Americans, and to a lesser extent Europeans, have been consuming in increasing amounts food products containing GM ingredients for several years without any apparent undue effects. The FDA would claim this as a justification of their approach to safety assessment. Some consumer-interest groups, however, argue that this is using humans as guinea pigs and that it is too soon to judge any long-term consequences.

Although maize and soybean represent the bulk of GM material entering the food chain directly or indirectly via animal feed, various other GM food crops have been grown in lesser amounts, principally in the USA and Canada. These include GM tomatoes, squash and potatoes. While fresh GM tomatoes have been sold in the USA (the so-called Flavr Savr tomato), an equivalent product developed in the UK was sold only as the puree and made solely from imported GM tomatoes and labelled as such. Since its introduction in 1996, over two million cans of this heat-treated product were sold before being withdrawn in 1999 by the producers in response to the general flight of UK supermarkets from GM produce.

Rapeseed (canola) is also a likely source of GM material. Although most rapeseed oil is of European origin and from non-GM varieties, some seed is imported from North America for sowing and for use in oil extraction. Most comes from Canada, where use of GM oilseed rape is commonplace (62% of the total crop in 1999). Imports into Europe of non-GM hybrid canola seeds from Canada were discovered in 2000 to contain some GM material (<1%). In the view of some member states, this represented an unapproved release of a GM crop to the environment and several thousand hectares of rape, principally in the UK but also in Sweden and Germany, were ordered to be destroyed. It is unclear whether this mixing of GM with conventional seeds occurred at the seed station or whether there was cross-contamination in the field. However, with the level of use of GM technology in North America, strict segregation will be virtually impossible to maintain and some adventitious contamination of conventional crops is inevitable. Studies originating in the USA have already detected traces of GM material in the majority of consignments of conventional maize tested. If European governments continue a policy of no latitude and to require crops containing trace amounts of GM material to be destroyed, sourcing of seeds from North America will become increasingly impractical.

In addition to ingredients, various additives and processing aids are also derived from GM organisms, most notably some enzymes used in food production. Virtually all hard cheeses are now produced with the aid of a chymosin, an enzyme of bovine origin, cloned into and produced by a bacterium (Escherichia coli), a yeast (Kluyveromyces lactis) or a filamentous fungus (Aspergillus niger). These three sources of nature-identical enzyme replaced the use of bovine rennet and were welcomed by the Vegetarian Society.

Genetic Modifications to Food Plants

It is easy to forget that, even in the USA, the first GM crops were only authorized for general release in 1993 and grown commercially in 1995, although such releases were preceded by several years of field trials. Regulatory authorities have had only around 10 years and a handful of practical examples with which to develop a safety assessment strategy. To date, less than 20 structural genes have been introduced singly or in combination into GM plants for which approval is sought for sale or growth in the EU, or whose products can be found in European food. The US experience would extend to perhaps twice this number of genes. Structural genes are also accompanied by regulatory sequences, which direct the expression of the structural gene in the plant. Although most attention has been paid to the safety of the products of structural genes, it has been suggested that some regulatory sequences of viral origin may also pose hazards, although most dispute this.

The structural genes to be found in transgenic plants already authorized for release in at least one country can conveniently be considered in six groups according to their function: insect resistance, herbicide tolerance, virus resistance, male sterility, gene silencing and selection markers.

Insect Resistance. Insect resistance is conferred by genes coding for a truncated form of the protein endotoxins produced by strains of a soil bacterium, Bacillus thuringiensis (Bt). Many forms of these toxins exist in nature, affecting different insect species. Maize and tomato crops engineered for insect resistance carry genes of the cry1(A) or cry9(C) type whose products are specifically toxic to lepidoptera like the European corn borer, while, in potato, incorporation of cryIII(A) provides protection against the Colorado beetle. Suspensions of bacteria producing Cry1A toxins have been used as crop protection biopesticide sprays for nearly 40 years, for both glasshouse and field application. Use of this spray is permitted for organic farmers who now are particularly concerned about resistance developing as a result of increasingly wide-scale and continuous use. However, Bacillus thuringiensis is a close relative of Bacillus cereus, the human enteropathogen, and produces the same human enterotoxins in addition to its insecticidal properties, making its continuing use as a biopesticide open to question.

Herbicide Tolerance. Crops expressing tolerance of one of two herbicides, glufosinate ammonium (Basta) and glyphosate (RoundUp), are the most commonly encountered. Examples of both have sought release in Europe. Elsewhere, plants expressing tolerance of bromoxynil, sulphonylurea and imidazolinone have received market approval. Glyphosate acts by inhibiting an enzyme involved in the synthesis of aromatic amino acids in the plant. Resistance is provided by a bacterial gene coding for the same enzyme which is sufficiently similar to replace the action of the plant enzyme allowing the production of amino acids, but sufficiently different to resist the effect of glyphosate. A maize line made resistant to glyphosate because of a deliberate mutation to the plant enzyme has also been developed and marketed. Glufosinate ammonium is an analogue of L-glutamic acid and a potent inhibitor of glutamine synthase. In most cases, resistance is provided by the introduction of genes (bar or pat1) encoding phosphinothricin acetyl transferases which catalyse the acetylation of glufosinate ammonium, rendering the compound inactive.

Virus Resistance. It has been known for many years that inoculation of plants with a mild strain of a virus can prevent infection by more virulent strains. Similar effects can be obtained by introducing genes coding for the virus coat protein alone. This was first demonstrated for the tobacco mosaic virus (TMV) in laboratory experiments and subsequently in field trials in 1988 made with tomatoes modified to express TMV coat protein. Virus-resistant varieties of squash, papaya and potatoes expressing the coat proteins of a number of different viruses or other viral proteins (replicase genes) now have market approval in North America. None, however, have sought approval in Europe to date.

Male Sterility. A variety of mechanisms have been developed to introduce male sterility. This has been dubbed ‘terminator technology’ because, in some forms under development, genes are included that act as molecular switches able to restore fertility, thus enabling companies to protect their intellectual property rights (see page 16). Considerable concern has been voiced about the impact of this technology on the ability of farmers in poorer regions of the world to save seed for replanting. In practice and to date, the technology has found only benign application in preventing self-fertilization and ensuring uniform hybrid seed production. In these cases, male sterility is introduced by barnase, a gene that codes for a ribonuclease and is controlled by a regulatory sequence that allows expression only in pollen. The presence of barnase renders the pollen sterile and prevents fertilization.

Gene Silencing. In addition to introducing genes expressing novel traits, recombinant technology can be used to inhibit or ‘silence’ the expression of existing, undesirable plant genes. This is always based on the introduction of an extra copy of the target plant gene, either in reverse orientation (antisense) or in the same orientation but in a truncated form (sense). The presence of the additional copy blocks the normal process of gene expression and leads to the suppression of the target enzyme. Probably the best known example is the tomato with increased self-life, in which the tomato enzyme polygalacturonase partly responsible for softening is inhibited. In the case of the Flavr Savr tomato introduced in the USA, the antisense gene was introduced. The risk of patent infringement prevented the UK producer taking the same approach, and so a truncated sense sequence was used for their product. Other applications with, or seeking, market approval are potatoes for starch production in which the granule-bound starch synthase is inhibited, producing starch with different properties of value to industry, and high oleic acid soybean in which a desaturase gene is silenced.

Selection Markers. The presence of antibiotic resistance markers in GM constructs has received a great deal of attention because of the concern about the growing resistance of pathogenic bacteria to the antibiotics used in their control. In reality, antibiotic resistance markers in plants do not pose any serious issues of safety. This is partly because the force of public concern already has ensured that alternative methods will be used to select for the introduced traits in GM modified plants in the future (see page 14). In the interim, the two most commonly used selection markers, nptII and aad A, which code for kanomycin/neomycin and streptomycin resistance, respectively, are antibiotics of very limited importance in human medicine. The remote possibility that either of these resistance genes could transform and be expressed in bacteria found in the digestive tract is of no significance against the background of resistance already present. Use of other antibiotic resistance markers such as ampicillin resistance (bla) is less desirable and, however slight, the risk associated with the introduction of amikacin resistance (nptIII), a reserve antibiotic of some importance, is unnecessary and unacceptable.

(Continues…)Excerpted from Food Safety and Food Quality by R.E. Hester, R. M. Harrison. Copyright © 2001 The Royal Society of Chemistry. Excerpted by permission of The Royal Society of Chemistry.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.