This section deals with specific issues that are frequently raised with regard to the safety of genetically modified foods. These issues include phenotypic analysis, sampling, compositional analysis and post-market surveillance. The discussion which follows provides an evaluation of existing knowledge about these topics and elaborates scientific approaches that may be used to assess possible health risks.
The selection process of the initial founders will be very limited compared with the plant breeding situation where thousands of GM calluses are screened for incorporation of the transgenic fragment and subsequently monitored for their phenotypic characteristics. This means that the information on the variation range between animals with the same genetic modification will be rather limited and that detected differences between individual animals will be difficult to interpret.
6.2.1 Phenotypic characteristics
Phenotypic analysis will relate to compositional analysis, but also to general performance parameters (such as growth rate, feed conversion efficiency, reproduction, clinical parameters), disease resistance and, depending on the animal and the genetic modification, the colonization and shedding of pathogens that may have food safety implications.
In specific cases, phenotypic analysis may also be advisable after processing or, for fish, during the various stages of spoilage. For example, adverse biogenic amines can be formed during spoilage in salmon, tuna, herring and other fish species. Similarly, formaldehyde may be formed in spoiled shrimp, cod, hake and many other species.
6.2.2 Sampling
For individual animal species the number of GM animals and conventional counterparts required for a compositional analysis needs to be determined in order to result in statistically reliable results, i.e. perform a power analysis. Furthermore it will need to be decided which edible tissues and products should be analysed in the different animal species.
It is clear that, in the case of GM fish and other aquatic species, more individuals are likely to be available for a compositional analysis compared with, for instance, large farm animals. In order to obtain statistically meaningful results in both cases it will be necessary:
to have information on the natural variation in the different tissues;
to apply standardized experimental conditions to the GM animals and conventional counterparts; and
to have standardized conditions for harvesting tissues and other animal products, for instance, with respect to developmental stage, age or market weight.
An example is the GM fish that has incorporated a growth hormone gene into its genome. In this case, sampling for the compositional analysis may be based on market weight rather than age.
6.2.3 Compositional analysis
For GM animal-derived food products the same basic approach for the compositional analysis should apply as for plants. Similar to the GM plants, the key constituents of the tissue would have to be established. This will include key nutrients as well as those compounds that may have adverse effects on human health, such as thiaminase in fish (Kleter and Kuiper, 2002) and wax esters in butterfish (Nichols, Mooney and Elliot, 2001). The list of key constituents per tissue will have to be flexible and may have to be adjusted based on the current state of the art in an ongoing process.
It is important to generate background data on the natural variation for the individual constituents in different tissues. Existing databases on the composition of animal food products need to be evaluated as to whether their data are of sufficient quality to be of value in a comparative compositional analysis.
Post-market surveillance requires food intake data and may require the establishment of adequate product tracing systems. In this respect the food animal sector has advantages over the crop plant sector where basic product tracing systems are still virtually lacking. In the animal production sector, such systems are already well established for some animal food production chains in some countries and many other initiatives are ongoing in this field. It will, however, require further elaboration and adjustment before these systems can be used for the purpose of post-market surveillance.
As a result of ongoing genome programmes, the potential of genetic modification in animal production will increase. The variety of genes that can be used will increase and it may also become more feasible to transfer entire metabolic pathways. It may be anticipated that the number of GM animal-derived products with improved health characteristics may increase, for instance pigs with a nutritionally improved meat composition or shrimp with reduced allergenic potential.
Based on our increased knowledge of the physiology of food animals, the number of unpredictable secondary effects of the genetic modification that cannot be analysed in a targeted analysis may decrease, theoretically reducing the number of unexpected effects of a transformation event.
Further development and validation of profiling methodologies in the field of genomics, transcriptomics, proteomics and metabolomics may provide additional insight into the unintended effects of a transformation event.
Improved product tracing and information transfer systems may make the application of post-market surveillance systems more feasible to assess the long-term effects of complex food products, including food products derived from GM animals.
