ACRE Advice

ACRE's response to concerns raised in written representations and submissions associated with the CHARDON LL public hearing and to statements made at ACRE's open hearing relating to the safety assessment of T25 GM maize conducted under Directive 90/220/EEC

Environmental Issues

Point 1. It has been asserted at the hearings that ACRE did not consider the effects on soil ecology resulting from the horizontal gene transfer of herbicide tolerance and antibiotic resistance transgenes from CHARDON LL to soil bacteria.

Part of ACRE's risk assessment for any GMO being considered for release into the environment is its potential impact on the soil ecosystem. For T25 maize, this includes the likelihood of transgenic DNA from this GM plant line (especially from its detritus) being taken up by soil bacteria and the consequences of this for soil ecology.

In the laboratory, conditions can be applied to force horizontal gene transfer between plant and bacterial cells, albeit at a very low frequency. However, there is no convincing evidence that horizontal gene transfer between plants and bacteria takes place in the soil under natural conditions. Two published studies designed to measure horizontal gene transfer between plants and microbes in the soil have been quoted as demonstrating that this phenomenon occurs ^14. However, the authors of these publications did not demonstrate, nor did they claim to provide evidence of, horizontal gene transfer between plants and bacteria under laboratory conditions or in field studies (see Point 2 for more details). This is not to say that that horizontal gene transfer between plants and bacteria could not happen, but the evidence suggests that if it does, it is a very rare event.

Taking a precautionary approach, ACRE assumes that horizontal gene transfer could occur between T25 maize and soil bacteria and then considers the implications. Horizontal gene transfer can only impact on soil microflora if the pat gene is expressed following transfer and if this expression confers a selective advantage to the recipient bacterium.

The cauliflower mosaic virus (CaMV) 35S promoter drives pat gene expression in T25 maize. If this promoter was transferred into a bacterial genome ACRE's view is that its capacity to regulate gene expression would be negligible. ACRE accepts there may be very low levels of activity, but would emphasise that the expression of genes regulated by the CaMV 35S promoter in bacteria is minimal.

The amount of PAT protein that could potentially be synthesised is also affected by the ability of a bacterium to translate any pat gene expression into protein. The pat gene in T25 maize has been modified to optimise the synthesis of the PAT protein in plants and to minimise its translation in bacteria. To achieve efficient expression of bacterial genes in plants it is common for researchers to modify the DNA codon usage pattern by altering the ratio of C/G base pairs to A/T base pairs. This does not result in an alteration to the sequence of amino acids that make up the final protein. As the pat gene isolated from Streptomyces spp. has a relatively high content of G/C base pairs when compared to plant genes, the ratio of base pairs was modified prior to introduction into maize. As a result, the modified gene is expected to lead to significantly less PAT protein in the bacterium that acquires it compared to the naturally occurring form of this gene.

ACRE also assessed the possible consequences to soil microflora if horizontal transfer of the transgene did occur and if it was then translated into an active protein. An important consideration in the risk assessment is whether L-PPT resistance is a trait that already exists in the UK's soil microbes. The fact that the particular L-PPT resistance gene used to transform T25 maize is specific to a laboratory strain of Streptomyces viridochromogenes (Wohlleben et al. 1988 ¹⁵)not present in UK soil bacteria was raised at the ACRE hearing. However, there are species of Streptomyces common to UK soils that do have L-PPT resistance genes. These include the bar genes of Streptomyces hygroscopicus (Thompson et. al. 1987 ¹⁶) and Streptomyces coelicolor (Bedford et. al. 1991 ¹⁷). Results from preliminary studies carried out by Dr. Jeremy Sweet at NIAB were quoted at ACRE's hearing as evidence for the lack of existing L-PPT resistance genes in UK soils. We understand from Dr. Sweet that he believed that his failure to amplify bar gene DNA to detectable levels was due to problems with the method used and could not be attributed to a lack of bar gene DNA in his soil samples.

Having established that L-PPT resistance genes exist in UK soil microbes, ACRE then sought evidence for bar/ pat gene expression in UK soil or its counterparts. There are relatively few studies on the occurrence of PAT activity in soil bacteria. However, Bartsch and Tebbe (1989) ¹⁸ showed that of the 300 diverse bacteria isolated from German soil samples, 2% exhibited PAT activity. Species of bacteria shown to have PAT activity in this research are also common in UK soils (eg. Rhodococcus, Alcaligenes, Agrobacterium, Serratia and Pseudomonas). In addition, this study identified enzyme activities that catalyse the degradation of the herbicide in a different way from the PAT protein.

One of the characteristics that led AgrEvo to select the T25 maize line for commercialisation was that it does not have an intact ampicillin resistance gene ¹⁹. Therefore horizontal gene transfer of antibiotic resistance is not an issue with T25 maize.

Taking all the evidence summarised above, the Committee's view is that the likelihood of genes moving from plant debris to soil bacteria is very low and that the environmental consequences (if it did occur) with this particular GM maize would not be significant. It is noteworthy that genes conferring PAT tolerance are already present in soil bacteria and it is much more likely that they will be transferred between bacteria by horizontal gene transfer than between plants and bacteria. Horizontal gene transfer is a well documented mechanism by which bacteria exchange genetic material in the environment.

Point 2. There has been concern that research by Gebhard and Smalla (1998 and 1999) ²⁰ shows that horizontal gene transfer can take place between plants and bacteria in the soil.

In July 2000 ACRE reviewed these papers by Gebhard and Smalla (1998 and 1999) and concluded that they provide important information and confirm advice given previously ²¹. The authors of these publications did not demonstrate, nor did they claim to provide evidence of horizontal gene transfer under laboratory conditions or in field studies. Horizontal gene transfer between plant and bacterial cells can be achieved by using forced conditions in the laboratory and by supplying pressure to select for the gene product and consequently the transgene itself. Although there is no evidence that this process actually exists under natural conditions it remains a theoretical possibility. Therefore, taking a precautionary approach, ACRE's safety assessments assume that it might occur. (see Point 1 above).

Point 3. Concern has been expressed that ACRE did not modify its advice in response to a report by the National Pollen Research Unit ²².

In 1999, ACRE reviewed the report published by The National Pollen Research Unit on maize pollen dispersal to determine if it raised any evidence not available when ACRE advised, in 1998, on the cross-pollination between an organic sweetcorn crop and a GM fodder maize in National List trials at a site in Devon ²³. Specifically, ACRE considered whether its previous advice needed updating in the light of this National Pollen Research Unit report ²⁴.

The Committee agreed with the main conclusions of the report i.e. that maize pollen can be carried by wind and insects over long distances, but found that the report did not provide any new information on cross pollination/ hybridisation frequencies. The report from the National Pollen Research Unit considered pollen as particles and did not take into account viability. Most pollen from Poaceae (the grass family of which maize is a member) has a short viability, which restricts the opportunities for gene flow. ACRE uses information based on hybridisation frequencies as this provides a much more accurate prediction of gene flow. These data derive from years of international experience growing conventional crops for seed production as well as data from a wealth of scientific studies measuring gene flow.

With respect to the GM maize grown in Devon in 1998, ACRE advised specifically on this case and took into account local topography, size and number of plots (pollen dilution was a key issue as only 6 plots out of approximately 1800 were GM at the Dartington site). It also took into consideration cross-pollination rates for maize, which is an internationally recognised criterion for ensuring high purity in seed production (UK Seeds Regulations, EC Seeds Directive and OECD Maize Seed Scheme). At a separation distance of 200 metres the Committee concluded that cross-pollination events between the organic crop and its GM counterpart would account for no more than 1 hybrid sweetcorn kernel in every 40,000.

In contrast, the National Pollen Research Unit's report on pollen dispersal proposed that 1 kernel in 93 could be a hybrid at a separation distance of 200 metres. The National Pollen Research Unit's calculation of hybridisation frequencies is an approximation based on mathematical modelling. The report stated as much and qualified the estimate by stating that a whole variety of environmental factors could vary the accuracy of this estimate such as local topography, field size etc.

Whilst ACRE considers modelling experiments on pollen flow interesting and useful, data from hybridisation studies used in conjunction with local details are much more accurate in predicating gene flow. ACRE therefore concludes that the National Pollen Research Unit's report on pollen dispersal does not alter its previous advice.

A number of other reports associated with gene flow from maize have been submitted to the CHARDON LL hearing. ACRE is aware of these and the local and seasonal variation that their results imply and it is content that the separation distances employed at the Dartington trial site minimised the probability of T25 maize hybridising with non-GM lines. It is important to note that separation distances are pragmatic - they are based on years of international experience producing high purity seed as well as from numerous scientific studies measuring hybridisation frequencies. They do not and were never intended to provide complete genetic isolation and as such GM crop varieties are only released into the environment if the consequences of gene flow are assessed as not presenting a greater risk to the environment or human health compared to their non-GM counterparts.

Point 4. Concern has been expressed that ACRE did not evaluate the safety of CHARDON LL in accordance with the more stringent amendments to Directive 90/220/EEC and the EU environment Minister's declaration of December 1998.

Until December 1998, the EU competent authorities had not agreed a harmonised approach to risk assessment. It was therefore important that some clear principles reflecting best practice in EU member states were formalised. In 1998, negotiations to amend the existing directive, 90/220/EEC started. In October 2002, Directive 90/220/EU was repealed and replaced by Directive 2001/18/EU. This new directive contains a number of procedural changes and amongst these are two new annexes, one concerns the principles for risk assessment (Directive 2001/18/EU: Annex 2) and another on post-market monitoring (Directive 2001/18/EU: Annex 7).

The new procedures reflect the issues that member states have considered during their safety assessments since Directive 90/220/EEC came into force in 1992. A fundamental property of the directive is that each member state makes an independent assessment of every application submitted to place a GMO on the market and these views are then shared before a final decision is made. A diversity of views and concerns are therefore considered for each application before member states collectively decide whether or not to grant a marketing consent.

With respect to Annex II of the new directive and its reference to direct, indirect, immediate and delayed effects - ACRE has always considered these aspects in risk assessments and this is evident from its minutes and advice. For example, ACRE's consideration of non-target/indirect effects includes any possible consequences resulting from the insertion of transgenic DNA into a host's genome. The transfer of transgenes to wild relatives and any environmental consequences that result, are a further example of ACRE's consideration of indirect effects. This is also an example of a delayed effect since environmental impact must be viewed over several generations.

It was possible under the original directive (90/220/EU) for member states to agree to impose post market monitoring. For maize varieties such as CHARDON LL, ACRE was satisfied that there was no evidence to suggest they would behave any differently from conventional varieties. On this basis, ACRE was confident that the maize did not need to be monitored over and above the normal quality control and inspection tests that are observed for all plant varieties. When the decision to issue a consent to place T25 maize on the market place was made in April 1998, no member state called for post-market monitoring.

The UK Farm Scale Evaluation (FSE) programme ²⁵, designed to assess the potential impact of farm management practices associated with the cultivation of herbicide-tolerant GM crops, has not yet been completed. No commercial planting of T25 maize will take place until the FSEs have concluded in 2003 and only then, if the use of L-PPT herbicides on maize is considered to cause no unacceptable effects on the environment. Approvals for the use of the herbicide on maize under pesticide regulations and the addition of T25 maize to the National Seed List are also required before it can be grown commercially in the UK.

ACRE's risk assessment procedures have been in line with the new directive for a number of years; indeed, UK Government scientists and ACRE have been pivotal in guiding the revisions for products considered at EU level under part C applications.

Point 5. It has been asserted that ACRE did not consider the possibility of genetic instability in T25 maize.

ACRE is satisfied that the transgenic material is inserted stably and there is no evidence that T25 maize is any less stable than its conventionally bred, non-GM counterparts. This assessment includes a consideration of transposable element activity ²⁶ and changes in the methylation state of DNA ²⁷. The original T25 maize line was created in the late 1980s. It was selected from a large number of transformed lines based on its molecular characteristics i.e. it has a single, relatively short insert with an incomplete ampicillin resistance gene. Subsequently, the transgene has been expressed under different genetic and environmental conditions through numerous generations without any evidence of instability ^28; in each case the pat gene conferred the expected tolerance to the herbicide.

During ACRE's hearing it was implied that T25 maize had failed the original distinctness, uniformity and stability (DUS) tests (criteria on which proposals to add varieties to the National List are partially based) on safety grounds - this is not the case ⁵. There is also no evidence in the literature to support the idea that transgenic DNA is inherently less stable than native DNA. It was pointed out by one of the expert witness at ACRE's hearing that all genomes are unstable to a certain degree and that there is no particular reason why transgenic DNA should be less stable unless it carries with it mechanisms for insertion and excision. As previously discussed there is no evidence in successive generations of T25 maize lines that the inserted sequence and therefore any genetic elements contained within it, confer instability.

Point 6a. It has been suggested that the possibility of T25 maize being toxic to bees has not been considered.

The application dossier submitted by AgrEvo in 1995, did consider the possibility that T25 maize could be toxic to beneficial insects. ACRE assessed the impact that T25 maize might have on beneficial insects such as bees taking into account their exposure to the transgenic product and consequently what effect the PAT protein might have. It is important to note that the genes themselves are not toxic; it is only the gene products (i.e. proteins) that have the potential to be toxic. Therefore, one consideration is whether the transgene is translated into protein (i.e. expressed) in particular tissues or structures, such as pollen.

T25 maize has been grown commercially as well as in research and development trials in Europe and the United States for many years and there is no evidence that it causes harm to bees.

• ACRE considers that T25 maize poses a negligible risk to bees since:
• There is no evidence that PAT is toxic to bees.
• Tests show that PAT enzymic activity is very specific for its substrate, PPT and it is only species within the genera Streptomyces and Kitasatosporia thathave been reported to synthesise this amino acid.
• Analyses show that PAT protein and enzymatic activity is absent in CHARDON LL pollen

Point6b. Concern has been expressed that bees could be at risk from horizontal gene transfer of the pat gene.

It has been suggested from preliminary studies by Dr Hans-Hinrich Kaatz ²⁹ that the pat gene from T25 maize could be transferred to bacteria in the intestinal tracts of bees exposed to its pollen. Since this work has not been published, ACRE cannot give its view. However, in line with the precautionary principle, ACRE does consider the possibility of horizontal gene transfer of transgenes to microorganisms and there is no evidence in this case, that it would have an adverse effect.

Point 7. There has been some concern that dust and sap from T25 maize plants might adversely affect humans and bees respectively.

Chemical analyses demonstrate that T25 maize is compositionally equivalent (see Point 12) to its non-GM counterparts except for the presence of PAT protein. All the available evidence supports the conclusion that this protein does not confer increased toxicity (e.g. toxicity studies and biochemical properties) or allergenicity (the PAT protein does not have significant homology to any known allergen) on T25 maize compared to non-GM maize varieties. Therefore, there is no evidence for, or reason to expect that sap, crop dust or any material derived from T25 maize plants presents an increased risk to human health or the environment compared to matter from conventional maize varieties.

Point 8. It has been asserted that the possible production of unidentified biochemical products in maize crops and wild relatives resulting from cross pollination with CHARDON LL was not considered.

As there are no sexually compatible wild relatives of maize in Europe, transfer of genes via cross-pollination events can only occur between CHARDON LL and other maize varieties. If CHARDON LL did hybridise with a different maize line the transgenic DNA would still be in a very similar genetic background. Since the transgene does not result in the production of unexpected toxins or allergens in CHARDON LL there is no reason for this to be any different in a hybrid. In addition, there is no evidence from a detailed knowledge of the mode of action of PAT that it would alter the metabolism of maize hybrids any differently from CHARDON LL.

The PAT enzyme has been well-characterised biochemically ^30; it is highly specific for its substrate, L-PPT (an unusual amino acid) and is incapable of catalysing reactions involving other closely related compounds (including other amino acids). This indicates that PAT is very unlikely to interfere with other plant processes.

Point 9. There has been a suggestion that the cauliflower mosaic virus promoter is unsafe and that its use can lead to gene silencing and DNA rearrangements.

The expression of the pat gene in T25 maize is regulated by the cauliflower mosaic virus (CaMV) 35S promoter. ACRE has considered the CaMV 35S promoter on a number of occasions and in particular reviewed the paper by Ho et al., (1999) ³¹. The Ho et al. paper discusses the potential for the CaMV 35S promoter to recombine with other virus DNA and the role that this promoter might play in horizontal gene transfer. The paper reviews the scientific literature on the CaMV 35S promoter and advances a hypothesis that a 'recombination hotspot' predisposes the promoter to recombine with virus, plant and animal DNA that has a similar nucleotide sequence. Ho et al. suggest that this may lead to the generation of new pathogenic viruses or to the unexpected and harmful over-expression of plant genes located in close proximity to the site of recombination. It was also suggested that recombination events with animal DNA might even lead to cancer. However, no new data or direct experimental evidence is presented to support the authors' hypothesis that the CaMV 35S promoter, used in many GM plants, is inherently dangerous.

For many thousands of years CaMV and its relatives have infected plants; consequently humans and animals have been eating plant material containing the 35S promoter via natural CaMV infection. No ill effects due to the activity or recombination of the virus promoter have been reported and in particular, no reports of cancer. In fact, Brassica crops such as broccoli, which usually carry at least some CaMV infection, are implicated in protection from cancer.

Ho et al., speculate that the 35S promoter has a tendency to recombine with other DNA which could have harmful consequences. However, there is no evidence for this, or any reason why 35S CaMV DNA would be more prone to recombination than other DNA that does not have elements associated with insertion and excision.

A prominent concern raised in this paper is that the 35S promoter in GM plants could inadvertently activate dormant viruses or non-target genes in plants or other organisms. ACRE is aware that the 35S promoter has the potential to alter the expression of host genes neighbouring the site of its insertion. This is one of the reasons why ACRE require applications for marketing consents to describe the host DNA that flanks the site into which the transgene has inserted. In addition, phenotypic and compositional studies are used to determine whether the direct or indirect effects of an insertion event alter important characteristics of the transgenic line. There are no reported incidences of a dormant plant virus being unintentionally activated by the insertion of a transgene with a 35S promoter. In particular, there is no evidence that the T25 insertion event has altered the maize line in any way that makes it less safe to human health or the environment than its conventional counterparts.

Ho et al., have suggested that there is a 'close relationship' between CaMV and human viruses such as hepatitis B and that CaMV 35S promoter DNA inserted into transgenic plants will recombine with DNA from these viruses. However, CaMV and human retroviruses are not members of the same genetic family and the degree of similarity between their DNA sequences is low. It is also worth considering that if recombination between plant and animal viruses, perhaps in the gastrointestinal tract, were a realistic possibility then the abundance of plant virus DNA in the diet would mean that this was a frequent natural occurrence and it is not.

In conclusion, the paper by Ho et al., does not challenge current scientific understanding or indicate that the CaMV 35S promoter, as used in genetic modification, is inherently unsafe. The CaMV 35S promoter used in this way represents an extremely low risk to human health and the environment.

14 Frank Gebhard & Kornelia Smalla (1998). Transformation of Acinetobacter sp. Strain BD413 by transgenic sugar beet DNA. Applied and Environmental Microbiology 64 (4) 1550-1554.
Frank Gebhard & Kornelia Smalla (1999). Monitoring field releases of genetically modified sugar beets for persistence of transgenic plant DNA and horizontal gene transfer. FEMS Microbiology Ecology 28, 261-272.

15 Wohlleben W., Arnold W,. Broer I., Hillemannn D,. Strauch E. and Puhler A. (1988). Nucleotide sequence of the phosphinothricin N-acetyltransferase from Streptomyces viridochromogenes Tu494 and its expression in Nicotiana tabacum. Gene 70: 25-37.

16 Thompson C. J., Movva N.R., Tizard R., Crameri R., Davies J.E., Lauwereys M. and Botternan J.(1987). Characterisation of the herbicide resistance gene bar from Streptomyces hygroscopicus.
EMBO J. 6: 2519-2523.

17 Bedford D. J., Lewis C.G. and Buttner M.J. (1991). Characterisation of a gene conferring bialaphos resistance in Streptomyces coelicolorA3(2). Gene 104: 39-45.

18 Bartsch K. and Tebbe C.C. (1989). Initial steps in the degradation of phosphinothricin (glufosinate) by soil bacteria. Applied and Environmental Microbiology 55 (3): 711-716.

19 Northern blots and enzyme assays indicate that the gene is not transcribed and that no antibiotic activity is evident in the maize tissue.

20 Frank Gebhard & Kornelia Smalla (1998). Transformation of Acinetobacter sp. Strain BD413 by transgenic sugar beet DNA. Applied and Environmental Microbiology 64 (4) 1550-1554.
Frank Gebhard & Kornelia Smalla (1999). Monitoring field releases of genetically modified sugar beets for persistence of transgenic plant DNA and horizontal gene transfer. FEMS Microbiology Ecology 28, 261-272.

21 ACRE Annual Report Number 7 (2000) ANNEX H. Advice to the Secretary of State on Horizontal gene transfer: genetically modified crops and soil bacteria. Available online: www.defra.gov.uk/environment/acre/index.htm.

22 Emberlin J., Adams-Groom, B. Tidmarsh, J. (1999) A Report on the Dispersal of Maize Pollen. National Pollen Research Unit, University College, Worcester. Report commission by and available from the Soil Association, Bristol House, 40-56 Victoria Street, Bristol, BS1 6BY

23 Advice for the secretary of State 23 June 1998. Genetically modified maize in national list trials adjacent to an organic farm in Devon. Available on ACRE's website: www.defra.gov.uk/environment/acre/pubs.htm

24 Advice for The Secretary of State 25 March 1999. A Report on The Dispersal of Maize Pollen Compiled by the National Pollen Research Unit and Commissioned by The Soil Association

25 Detailed papers on the history and science of the Farm Scale Evaluations are available: www.defra.gov.uk/environment/fse/index.htm

26 Transposable elements occur naturally in many plants, including maize; they can move at low frequency from one genetic location to another

27 The methylation status of DNA can affect the expression of genes

28 Bayer CropScience estimate 23 generations of breeding, with 40 different maize varieties world-wide now containing the T25 insert.

29 University of Jena and the Hans Knöll Institute for Natural Substance Research in Jena.

30 In the presence of acetyl-CoA as a co-substrate, PAT catalyses the acetylation of the free amino group of L-PPT to yield N-acetyl-L-PPT, a compound that does not inactivate glutamine synthetase. Both of the PAT enzymes are highly specific for L-PPT and do not acetylate other L-amino acids, nor do they acetylate D-PPT. In the presence of excess concentrations of L-amino acids, both PATs also are unaffected in their ability to acetylate L-PPT. In L-PPT tolerant plants which express relatively high levels of PAT, the main residue metabolite of L-PPT catabolism is N-acetyl-phosphinothricin

31 Ho, M. W., Ryan, A., Cummins, J.(1999). Cauliflower mosaic viral promoter - a recipe for disaster. Microbial Ecology in Health and Disease 11 (4)