Results

Progress Working Group 2: Detection of Unintended Effects

The following issues were presented and discussed by the working group:

• Definitions of intended/unintended effects and targeted/non-targeted analysis
  

• Compilation of literature and data on parameter selection for intended/unintended effects and selection of comparators
  Information from this literature review and from sources such as WHO, OECD, FAO, will be used as a basis for detecting unintended effects.
  OECD consensus documents on potato, bread wheat, rapeseed and rice are available from http.//www.oecd.org/ehs/cd.htm.
  Issues at the molecular level were considered in detail in a working document. This document compared the integration of foreign genes into the host genome by procedures used in the deliberate genetic modification of plants (Agrobacterium-mediated transformation, or gene integration by physical techniques such as the biolistic method) with that of natural recombination of DNA in plants. To date, there is considerable data available on the integration of T-DNA (biological methodology), but not for vector DNA (physical methodology). At least for T-DNA, integration is preferential for regions which contain genes; this is also true for natural recombination of DNA in plants. Thus, in this respect, the element of risk of GM technology is no greater than that which is known to occur in nature. In respect of vector DNA integration, it was considered desirable that vectors should be designed whereby extraneous backbone sequences are removed.
  It should be made clear that an unintended effect does not imply that there is necessarily a risk. Information on insertion sites should help in anticipating any potential risks, thus providing the opportunity to eliminate them. Thus this type of information is one means of reducing the possibility of having unintended effects; in addition, it further contributes to the characterisation of the plant at the molecular level.
  Recommendations: Information on insertion sites should be made mandatory. Vectors lacking extraneous backbone sequences would be desirable.

• Unintended effects in conventional breeding
  It was described that owing to the fact that the occurrence of unintended changes is an inherent characteristic of conventional breeding processes, it is difficult to find references specifically describing such unintended effects and the way they are being handled. Data will be provided on the characterisation of conventional breeding on the DNA level. Breeders will be contacted for information on documented cases of unintended effects.

• Present status of approaches in demonstrating substantial equivalence/unintended effects

• Literature on successful application of targeted approach to GM commercial crops
  A working document was presented outlining the application of targeted approaches to demonstrate substantial equivalence. Such approaches include analysis of macro-nutrients, amino acids, fatty acids and nutrients such as vitamins and essential amino acids. Information and relevant references were provided giving specific examples carried out on GM crops to date. Information on the Safety Assessment of Novel Food Plants (Chemical analytical approaches to the establishment of substantial equivalence) has been published by The Nordic Council of Ministers (TemaNord 1998:591). The minimum lists for compositional analysis of specific crops recommended in this document were considered as useful guidance.

• Literature on observed unintended effects in transgenic crops using non-targeted approach
  The working document considered the technologies involved in plant transformation, and its contribution to plant breeding, including gene stacking. Further, the issue of unintended effects (impact of random insertions, impact of host factors) and non-targeted approaches to analyse unintended effects (genomics, proteomics, metabolomics) were considered.
  
  The potential of genomics, proteomics and metabolomics was discussed. After standardisation these different approaches will be valuable in the development of procedures for investigating unintended effects. The sequential use of these approaches compared to their simultaneous use was also discussed. Proteomics was considered to be an important tool in the initial safety screening procedure of GM crops. A schematic flow chart to compare the proteome profile of a given GM crop with the non GM parental line from which it was derived, was given. For metabolite profiling of plants 3 main methods are utilised, GC-MS, HPLC-MS (MS/MS), NMR; specific examples for potato were given (Roessner et al. Plant J. 2000).

• Biological and safety relevance
  Two issues concerning unintended effects have to be addressed: the biological relevance to the crop, and the safety to the consumer. Based on a knowledge of the crop and gene to be inserted it may be possible to predict, to a limited extent, both the intended and unintended effects.
  Clear statements on scope and limits of the emerging profiling techniques, especially the requirements to make the broad spectrum of generated data useful for a safety assessment, were considered necessary.

The following framework was agreed upon for the document to be prepared by the WG:

Detection of Unintended Effects - Present Status

1. Introduction
2. Definitions

• Intended and unintended effects
• Predictable and unpredictable unintentional effects
• Targeted and non-targeted analysis

3. Reasons for unintended effects

   Unintended effects in conventional breeding

4. General issues related to the detection of unintended effects

• Selection of appropriate parameters/comparators
• Natural variations (base line data)
• Statistical significance

5. Present status of approaches

Targeted approach

• Demonstration of "substantial equivalence" by investigation of defined constituents (1st generation GM crops)
• Demonstration of unintended effects by investigation of defined constituents (experimental crops)

Non-targeted approach

• Genomics
• Proteomics
• Metabolomics

6. Issues involved in using profiling techniques

• Standardisation/validation
• Data bases/bioinformatics

7. Biological and safety relevance

   (This aspect will overlap with the objectives of Working Group 1.)

8. Recommendations