The introduction of genetically modified organisms (GMO) in many countries follows strict regulations to ensure that only safety-tested products are marketed. Over the last few years, targeted approaches have been complemented by profiling methods to assess possible unintended effects of transformation. Here we used a commercial (Affymertix) microarray platform (i.e. allowing assessing the expression of approximately 1/3 of the genes of maize) to evaluate transcriptional differences between commercial MON810 GM maize and non-transgenic crops in real agricultural conditions, in a region where about 70% of the maize grown was MON810. To consider natural variation in gene expression in relation to biotech plants we took two common MON810/non-GM variety pairs as examples, and two farming practices (conventional and low-nitrogen fertilization). MON810 and comparable non-GM varieties grown in the field have very low numbers of sequences with differential expression, and their identity differs among varieties. Furthermore, we show that the differences between a given MON810 variety and the non-GM counterpart do not appear to depend to any major extent on the assayed cultural conditions, even though these differences may slightly vary between the conditions. In our study, natural variation explained most of the variability in gene expression among the samples. Up to 37.4% was dependent upon the variety (obtained by conventional breeding) and 31.9% a result of the fertilization treatment. In contrast, the MON810 GM character had a very minor effect (9.7%) on gene expression in the analyzed varieties and conditions, even though similar cryIA(b) expression levels were detected in the two MON810 varieties and nitrogen treatments. This indicates that transcriptional differences of conventionally-bred varieties and under different environmental conditions should be taken into account in safety assessment studies of GM plants.

Maize is a major food crop and genetically modified (GM) varieties represented 24% of the global production in 2007. Authorized GM organisms have been tested for human and environmental safety. We previously used microarrays to compare the transcriptome profiles of widely used commercial MON810 versus near-isogenic varieties and reported differential expression of a small set of sequences in leaves of in vitro cultured plants of AristisBt/Aristis and PR33P67/PR33P66 (Coll et al. 2008). Here we further assessed the significance of these differential expression patterns in plants grown in a real context, i.e. in the field. Most sequences that were differentially expressed in plants cultured in vitro had the same expression values in MON810 and comparable varieties when grown in the field; and no sequence was found to be differentially regulated in the two variety pairs grown in the field. The differential expression patterns observed between in vitro and field culture were similar between MON810 and comparable varieties, with higher divergence between the two conventional varieties. This further indicates that MON810 and comparable non-GM varieties are equivalent except for the introduced character.