The application of transcriptomics in the comparative safety assessment of (GMO-derived) plant products

National and international organizations have discussed current approaches to the safety assessment of complex (plant) food products in general and the safety assessment of GMO-derived food products in particular. One of the recommendations of different expert meetings was that the new analytical techniques, in particular the ‘omics’ approaches, need to be explored for their potential to improve the analysis and thereby the toxicological and nutritional assessment of complex (GMO-derived) plant products. This thesis aims to further explore this approach in general and, more specifically, has evaluated the potential added value of transcriptomics to assess unintended side effects in a newly bred (genetically modified) plant variety. As one of the first initiatives in this area a small food safety–related tomato-array was developed with pathway-selected cDNAs on the basis of two subtractive cDNA libraries. This tomato array was used to hybridise mRNA derived from tomatoes in five subsequent ripening stages from green, via breaker, turning, and light red, to red, to obtain a background library of gene expression profiles of different ripening stages for future comparisons. At the same time these initial series of experiments were aimed to assess the potential of the approach with respect to its sensitivity and specificity. In addition the tomato array was used to hybridise mRNA derived from GM tomato transformant lines and the traditionally bred parent line and the results were analysed for the presence of differential expression patterns in both transformant lines, that have incorporated the same genetic construct, compared to the parent line. A similar study was performed in Arabidopsis to assess the extent of unintended effects in GM lines that have incorporated different numbers of the introduced genetic construct. The resulting data show that the methodology of transcriptomics has the potential to detect large as well as small differences in gene expression. The first was primarily shown in the comparative study on the developmental stage, the latter in the comparison of transcriptomics profiles of the two transformant lines vs the parent variety. It was also shown that, for direct comparison, plants to be sampled need to be grown under very similar conditions and the sampling needs to be performed in a structured way taking into account the developmental stage of the selected plant organs and/or tissues. All experiments illustrated the necessity to establish the bandwidth of natural variation for comparative purposes in order to determine the biological as well as toxicological and/or nutritional significance of differences detected in GM lines or in lines resulting from other breeding procedures. Finally, the results of the international debate on the assessment of complex (GMO-derived) plant products, the knowledge on current breeding strategies, and the results of the first publications on experiments that aim to detect unintended effects in plant breeding strategies using ‘omics’ technologies, including the experiments described here, are combined to review current approaches. A new overall approach for the safety evaluation of complex plant products, including GMO-derived products, is proposed. This approach applies currently available tools, including the ‘omics’ technologies, to assess food safety aspects of newly developed plant varieties already during the plant breeding process. Undesired effects of the breeding procedures for the plant’s physiology, can thus be traced at an early stage and this will help to further guarantee the safety of the final plant-derived food products. Observed differences in the final plant product will form part of the comparative safety assessment. It can be envisioned on the basis of the data presented in this thesis as well as in other studies described in the scientific literature that in general few differences will be observed between GM lines and the WT counterparts that fall outside of the bandwidth of natural variation of commercial counterparts. The toxicological and nutritional evaluation of the final plant products will need to focus on the limited number of differences that are outside of this bandwidth and may affect the product’s food safety characteristics. To avoid inequalities in food safety assessment procedures that do not have a sound scientific basis, it is argued that these developments should have an impact on all novel plant varieties, not just on GMO-derived food plant products.