GM risk assessement

The latest UK John Innes Centre (JIC) annual report (1998/99) lists its staff publications in 1998. Two papers in particular deal with risk assessment in transgenic crops. Critics of GM technology are often accused of being irrational and emotive in their expressions of concern over the introduction of transgenic crops.

It is therefore interesting to read the extracts below from recent papers produced by leading scientists from the JIC on the subject of GM crop risk assessment. The JIC is one of the leading institutions responsible for the introduction of GM crops in the UK and a principal adviser to the UK Government on biotechnology.

The first paper by Drs Dale and Irwin is concerned with transgenic crops generally, whilst Professor Hull's paper is more specifically concerned with crops currently being introduced in some parts of the world which are genetically modified for protection against viruses. Professor Hull's paper is particularly concerned with the possible inadvertent creation of viruses with new characteristics through 'recombination' and other phenomena affecting transgenes in GM crops.

Both papers express concern in particular about biosafety issues that may not be detected in small scale trials, whilst offering little plausible suggestion as to how these might be dealt with in larger scale releases. In fact, Professor Hull suggests in his paper that this "will be much more difficult, if not impossible."

Several post-commercialisation effects not picked up during the testing and approvals process have in fact already been identified in a number of GM crops authorised for use in the United States (see http://www.btinternet.com/~nlpwessex/Documents/gmagric.htm ).

Interestingly much of the material in the JIC paper extracts below undermine claims often made by the biotechnology industry as to the 'precision' of genetic engineering (some elements of which are alluded to by the JIC authors). The acknowledgement by Dale and Irwin that in relation to the evolution of genetic balance in organisms the position of genes within the genome has 'adaptive significance' is particularly important. This is because with genetically modified crops the insertion of the transgenes and their ultimate positioning on the chromosome is a random process over which the 'engineer' has little or no control (even by the time of regulatory approval the position of the transgene on the plant genome may be unknown - see comment by Professor Bevan Mosely at http://www.btinternet.com/~nlpwessex/Documents/gmoquote.htm ).

These papers are also interesting in the context of claims (sometimes made by senior members of the scientific community) that 'genetic modification' is little different to conventional plant breeding. They also confirm that this 'science' has few predictive models concerning the behaviour of transgenes.

In the context of these papers, coming as they do from the very heart of the biotechnology scientific community, any claims that GM crops present no more of a risk to biosafety than conventional crops seem especially bogus. These factors are, of course, already well known to critics of the technology. However, it is nonetheless interesting to see them spelt out to a substantial degree in published papers by authors from a research institute which is jointly funded by the UK Government and the biotechnology industry, and which is well known for its pro-active promotion of the use of transgenics in agriculture.

Significantly, however, neither of these papers make explicit reference to the special risks associated with the use of the Cauliflower Mosaic Virus promoter used in most transgenic crops (although Dale and Irwin may refer to this by inference when stating that 'most gene constructs contain some sequences derived from plant pathogens'; and Hull may also refer to it by inference when alluding to 'interactions between the viral or virus-related sequences').

Abscence of explicit discussion of this especially critical area in the context of biosafety risk assessment is perhaps particularly notable given that the JIC has itself produced work (see previous JIC annual report) on this element relating to its influence on gene silencing (and more recently also on its relationship to recombination phenomena). For more information on the particular risks associated with the use of the Cauliflower Mosaic Virus promoter in transgenic crops see: http://www.btinternet.com/~nlpwessex/Documents/camv.htm .

Dale P.J., Irwin, J.A (1998) 'Environmental Impact of Transgenic Plants', Transgenic Plant Research (Ed:Lindsey K). Netherlands, Harwood Academic Publishers, 277 - 285

Because we rely largely on our knowledge of classical genetics to inform biosafety assessment, it is important to determine whether transgenes are the same as resident genes in every respect. As there are some differences, it is important to consider the consequences of those differences for influencing biosafety assessments. Some impacts are scale dependent and need to be examined carefully when transgenic plants are in widespread production......

We will consider issues that are scale dependent and likely only to be answerable following large scale, and even commercial production of transgenic crops.....

The essential difference between conventionally bred varieties and transgenic varieties is that the former are produced by cross pollination between varieties of the same species or between closely related species, while the latter can use genes from any class of organism. The consequence of this is that conventional breeding is limited to the genes (the gene pool) from the same species or closely related species; whereas for genetic modification, genes can be taken from viruses, bacteria, unrelated plants, animals (including humans) and made synthetically.....

By transformation it is possible to introduce genes into plants from any class of organism and therefore it is considered potentially possible to produce a plant phenotype of which there is little or no experience (Dale, 1993).....it is possible to change plant phenotype fundamentally by changing just one gene...

In the European Union (Directive 90/220; CEC1990b) all plants produced by transformation are assessed (process based assessment), whereas in the USA and Canada only transgenic plants modified with particular genes are regulated (product based assessment)..... As most gene constructs contain some sequences derived from plant pathogens or use the modified plant pathogen Agrobacterium for the transformation process, in practice virtually all of the products of transformation have so far been regulated in North America....

During evolution it is reasonable to assume that genomes have evolved a degree of genetic balance so that the position of genes within the genome has some adaptive significance. With transformation, it is possible to introduce alien gene sequences into random locations....

Although it is possible to study transgenes with a level of precision that is often difficult for resident genes (i.e. we know more about them) transgenes do often display unusually high levels of expression and structural instability....

The process of methylation is probably a mechanism that has evolved to neutralise invading DNA. The plant may therefore, in some circumstances, treat DNA inserted during the transformation process in a similar way to invading pathogenic nucleic acids. The difficulties that have been experienced in developing transformation methods in many species over the past decade of research, illustrate that plants do not readily take up and incorporate "invading DNA" into the genome.....

From Table 17.3. A comparison of the characteristics of resident genes and transgenes.

[With transgenic plants a] comparison of the expression of single copy transgenes from different transformation events (ie in different locations in the genome), frequently shows considerable variation in transgene expression.............

There are instances where the environment has significant effects on transgene expression .......

Transgenes have been demonstrated to influence characters apparently unrelated to the function of the transgene......

Transgenes transferred sexually into different genetic background have been shown to vary in expression, stability and pattern of inheritance......

Various kinds of interactions have been observed. Transformation involves the insertion of a novel piece of DNA , so there is no direct equivalent of the heterozygous condition. When the transgene is present on one homologous chromosome, it is in the hemizygous condition. Although not impossible (through chromosome deletions, translocations or transpositions) the hemizygous state is unusual for endogenous genes. Various kinds of inter-loci interactions involving transgenes have been observed including, co-suppression where a transgene can modify the action of an endogenous gene, and interaction between transgenes when one transgene construct can modify the action of another..............

Very high levels of transgene instability can be observed, especially among primary transformed plants, and early sexual generations. Instability can be of various kinds: transgenes or parts of them can be lost; expression can be silenced; transgenes can display abnormal inheritance; transgenes can show abnormal tissue specificity and expression can vary through plant development and inheritance....

Transgenes can interact with resident genes, and with other transgenes. Although there are undoubtedly complex interactions between resident genes, an understanding of the nature of interactions involving transgenes is important because of the likelihood that breeders will mix different constructs in plant varieties in the future (Flavell et al, 1996; Senior and Dale, 1996). There is also the possibility that transgenes will transfer by cross pollination from crop plants to wild and weedy species, and that wild species could receive several transgene constructs in natural and agricultural environments (Dale,1994; McPartlan and Dale, 1994; Scheffer and Dale, 1994; Scheffer et al., 1994, 1995)............

There is clear evidence that transgenes can be unstable and become silenced..... Transgene instability could obviously present a commercial risk to the plant breeder; but is transgene instability in itself a biosafety issue?......

Where transgenes are used to down regulate an allergen or potentially toxic substance in plants (e.g using co-suppression) silencing of the process may lead to activation of the formerly suppressed plant process...... It is possible in the future that pathways with significant environmental impact will be down regulated by transgenes and therefore the biosafety of instability in that modification would need to be considered.

Areas of sequence homology in transgene constructs are known to interact with one another in some plants. An effect of this can be that one transgene construct may inactivate the expression of another construct in the same plant. If these plants are reproduced sexually, it may be possible to separate the two constructs by segregation in subsequent generations. In this instance it is possible that a formerly silenced or poorly expressing construct may become more highly expressed in subsequent generations.....

When transgenic plants are evaluated, it is important to determine whether there is evidence of enhanced or modified transgene expression in different environments....

One of the most challenging issues associated with assessing the environmental impact of transgenic plants when in widespread commercial production are scale dependent effects. It is possible that a rare event may have insignificant consequences when transgenic crops are grown on a small experimental scale, but become more important when transgenic crops are grown over thousands of hectares.......

It is unrealistic to assume that every facet of environmental impact can be addressed and predicted from assessment of small scale experimental plots.....

Hull R. (1998) 'Detection of Risks Associated with Coat Protein Transgenics', Methods in Molecular Biology: Plant Virology Protocols: from Virus Isolation to Transgenic Resistance (Eds: Foster G.D., Taylor S.C.). New Jersey, Humana Press Inc. 81, 574-555

"Recent advances in the understanding of the molecular mechanisms of how viruses function and how they interact with plants have led to the development of various nonconventional approaches to protection of plants against viruses. Many of these approaches involve the introduction of viral or virus-based sequences into the plant's genome....

Many of these transgenic plant lines... have reached the stage of field testing for the efficacy of protection, and are even being more generally field-released. This raises the question of possible risks that could arise on general field release, a topic which has previously been discussed by Hull (5-7), Hull and Davis (12), de Zoeten (17), and Tepfer (18). Despite these discussions, the issue has not been fully resolved, and various other aspects are being raised.....

".......The area of concern specific to viral transgenes is the potential risks on any interactions between the viral or virus-related sequences being expressed from the transgene and another virus superinfecting that plant. Three main scenarios are usually considered: synergism, recombination and heteroencapsidation........

1.3. Recombination

Three sorts of recombination have been recognized (20): homologous with crossovers between related RNAs at precisely matched sites, aberrant homologous with crossovers between related RNAs not at corresponding sites, and nonhomologous with crossovers between unrelated RNAs at noncorresponding sites. There is considerable evidence for extensive recombination in RNA viruses (see refs. 20 and 21 for details), and probably all three mechanisms have been involved at one time or another. It is generally considered that recombination plays an important role in the evolution of RNA viruses (see refs. 20—23). Evidence is now forthcoming of recombination between superinfecting viral RNA and RNA expressed from a transgene (24) through the aberrant homologous recombination mechanism. The finding of recognizable host RNA sequences within viral RNAs (25,26) is suggestive of nonhomologous recombination.

All the experimentation on recombinants between plant virus sequences has been done in controlled laboratory situations. It is difficult to devise detailed protocols for the detection of recombinants produced in the field.....

This involves the superinfection of a plant expressing the CP of a virus, say virus A, by an unrelated virus B. Heteroencapsidation is the encapsidation of the genome of virus B by the CP of A, thereby conferring on virus B properties of virus A. There are several examples of heteroencapsidation in transgenic plants, both between viruses of the same group (27,28), and between unrelated viruses (29). The main property of CP that is considered is that of vector transmission characteristics. However, there is increasing evidence that CPs are involved in long distance viral movement around infected plants, and heteroencapsidation could enhance the movement of a superinfecting virus that did not normally move systemically (see Subheading 1.2.).

The discussion of heteroencapsidation has focused on superinfecting viruses. However, there is the possibility that heteroencapsidation of retrotransposons could present a problem. Retrotransposons are a major class of transposable elements whose structure resembles the integrated copies of retroviruses, and which are considered to be important in evolution (see ref. 30). The Tyl-copia group of retrotransposons is widespread in plant genomes (31—33), and it has been suggested that there might be horizontal transmission between species (31). Sequencing has shown that most copies of the Tyl-copia retrotransposons in plants are mutated, so they would not be active. However, several active ones capable of retrotransposition have been described (30,34—37) and presumably replicate, as do all retrotransposable elements, via RNA. Among the factors that activate plant retrotransposons is tissue-culture, a process involved in transformation (37). This raises the possibility that introduction of the CP transgene could activate retrotransposon RNA, which becomes heteroencapsidated and transmitted horizontally to other species.

The main question to be addressed is whether the risk on field release of the transgenic plant is significantly more than the risk from the nontransgenic plant ....

It is likely that it will take some time for a full risk assessment on the viral transgenic plants to be performed and commercial and other pressures will be very strong for field release. There are two approaches to risk reduction and control that can be put into effect relatively soon. One is to use biological containment ...... Much more difficult is to avoid recombination. ......

The second approach is to design methods for monitoring the effects of field release. For small-scale releases, it is relatively easy to design monitoring procedures for analyzing pollen flow into related weed species and for detecting heteroencapsidants or recombinants. This will be much more difficult, if not impossible, for large-scale releases, in which the approach should be to educate farmers and extension service personnel to identify any unusual event that might be associated with transgenic plants. This will be the challenge for the future.....